CEO of Semtech Corporation. We have a full day ahead of us. Before we start, let me just introduce you to the Semtech team so that you can interact with them during the day here. We'll start with Emeka. Emeka Chukwu is our Chief Financial Officer. We have Gary Beauchamp, who's our General Manager for our Signal Integrity Products Group. Mark Costello, who's our General Manager for our Protection Product Group. Marc Pégulu is our General Manager for our Wireless and Sensing Product Group. All three of those executives will be presenting today. In addition, we have Alistair Fulton, who's the head of marketing for our Wireless and Sensing Product Group. Madhu Rayabhari, who's at the back there, is the head of marketing strategy for our Protection Product Group. We have Bharat Tailor, also at the back there, who's our head of strategy for our Signal Integrity Products Group.
We have Chris Chang, who's our corporate business development and marketing executive. We have a couple of admin, Nicole Lucio and Sara Kesten, are helping us with the event today. Those are all the Semtech employees. We also have some demos in a room next door here. You'll see demos on protection, some demos on wireless and sensing, our LoRa products, and also some videos of our Signal Integrity Products. Please take the time and opportunity to take a look at those. There are some very interesting things we're trying to demonstrate there. We have a full agenda. I'm going to start with how the Semtech journey is going and where we are in that journey. I'll hand it over to Gary Beauchamp. Gary will talk about Signal Integrity Products, specifically how we're powering the data center revolution.
Mark Costello will then talk about protection products and the golden era of protection, why we believe this industry is about to really transform into a very exciting one for us. Then we'll spend quite some time on LoRa, specifically the Wireless and Sensing Product Group. Within that, we'll talk about LoRa and spend some time on how that's really transforming both Semtech and the industry in the low-power wide-area network space. Then Emeka will wrap us up with how the financial story comes together at the end of the day here. What are the messages for today and the goals for today? Before we really get into this, the message that I want to send to you, and I'll start with this and I'll end with this, is that in the history of Semtech's been around for over 60 years.
I've been the CEO for about 12 years now. Actually, over 12 years. There's no doubt in my mind, this is the best time in the company's history from many perspectives. Very exciting time for us. Very exciting time from a product standpoint, from an end market standpoint, from a customer standpoint, from an application standpoint, from a financial standpoint. As I look at it, where we are today, it's just a really exciting time for us. There's just many things going on that are exciting and very much. We've been implementing this strategy over many years, we're starting to see now the fruits of that strategy come together. That's essentially what we're going to explain to you today, is why we feel that way and some of the different elements of our business.
Some of the goals today are to introduce you to the strong leadership team. Many of you know me. I've had many discussions with you. Many of you know Emeka and Sandy, our investor relations director. To introduce you to the rest of the team, to get a perspective that this is really a true team operation here, and to really show you the bench strength that we have. We'll talk about our path to $1 billion in net sales and beyond, and why we think we can get there within a reasonable timeframe, and how we're going to drive growth beyond that. We'll talk about our growth engines, and then we'll talk specifically about some elements of the business that are non-financial, but elements that, since I've been with the company, I've said are very important.
Our diversification, our end market balance, our geographical balance, our portfolio balance, and our financial balance. You'll hear me talk a lot about balance, and I say that this balance word, we put this in place because I've been in this industry for over 30 years, and I've seen many cycles and I've seen many transitions within end markets, and I've seen many geographical issues, much like the trade issues we have today. Sometimes you have to sit back and say, "You know what? I want to mitigate against that." So the balance in all of these areas is what we try to achieve. I'll show you how we do that, and there's a reason behind it. Then finally, we'll talk about, in addition to the engines that are growing and driving our growth today, some of our future innovations, which are both unique and exciting.
It's really part of our culture. We focus a lot on innovation. When people ask me, when customers ask me, and partners ask me, "How do you keep your edge? How do you maintain when you've got competition coming in from all regions of the world? How do you stay ahead?" The key is innovation. There's no stopping people copying you or trying to rip your products off or take your technology and even steal it. The way to stay ahead is to continue to innovate. I mentioned our team and some of the executives I've introduced you to. I think it's a world-class team, it's very experienced, it's culturally diverse, and we all share the same values. 13 executives here, 11 of them are from different countries.
We have representation from Nigeria, from Switzerland, from Canada, from England, from France, from Israel, from Korea, Scotland, and Sri Lanka. We are truly a very diverse group. The one thing I would say is that this wasn't by design. We just chose the best executives, and we ended up with a very diverse team, a very strong team. Through our history, Semtech has grown through acquisitions and organically. A few points to mention, really, with our acquisition strategy, we don't do acquisition for scale. In general, we do acquisition that makes strategic rationale. I like to think about things that one plus one equal five. How do we make the target company and our company come together in a way that drives strategic value in the industry? Very different kind of philosophy than most companies view acquisition, but that's the way we've always done it.
We will do acquisition for SAM expansion. In the past, if the opportunity that we're faced with is not big enough, I may go to the board and say, "We need to grow our market opportunity, and therefore, we need to do an acquisition." In general, it's for SAM expansion or diversification or for strategic rationale. Through the years, I can give you some examples of that. Acquisition of Xemics gave us wireless transceivers. The acquisition of Gennum gave us an entry into the optical communications, and today is our data center plate. The acquisition of AptoVision gives us the entree into the Pro AV space. The acquisition of IC Interconnect, just recently, gives us the ability to accelerate some of our innovations and protection.
Throughout the years, we do acquisition, and we'll continue to use our balance sheet to help us in this area as we go forward. Over the last 12 years, in my journey here as the CEO, we have transformed the company. The portfolio has totally transformed. The end market have transformed. Even the infrastructure. Infrastructure, meaning our own internal ability to provide world-class operations, including the implementation of SAP as an ERP system, Workday, and Salesforce, which on the surface of it, you say, "Well, everybody does that." Turning all of those systems into enabling our ability to grow the company without growing headcount in a large way. Not the R&D side, but the SG&A side specifically, has been a key element of our strategy.
We feel really good about where we are from an infrastructure standpoint, and that's the ability to now drive the growth without adding a lot of OpEx, which Emeka will talk about. The other most really important thing here is the balance in end markets. You'll see our growth in enterprise computing space. I'll talk about this when I talk about end market balance. What I mean by that is when we see that too much of our business is exposed to one end market, like consumer, or too much is exposed to industrial, we think strategically about, okay, how do we rebalance? Part of this storyline is to try to make sure we do have that end market balance. The other part of the story is obviously growth. We've had some challenges.
Over the years, we've had some ups and downs, and sometimes you learn from these. Sometimes you have too much customer exposure, too much end market exposure. In the case of FY 2016, too much exposure to Huawei, too much exposure to Samsung, we try to offset that and fix some of that going forward. I think today, as I said, we're in the best position that the company's ever been in. One of the reasons I say that is that we feel that we have a lot more balance there in markets and customers. FY 2018 was a very special year for us. The way I like to think about the financial performance is the result of all the strategies you put in place and the execution. I felt, for the last several years, our strategy's been on point. We have a good strategy.
We have very good execution. We're starting to see the results. FY 2018 was the first example of that. We had record sales, record operating income, record earnings, record revenues for both our Signal Integrity Products Group and our Wireless and Sensing Product Group, record design wins. What's more important, though, than the financial performance of FY 2018 is the storyline underneath it that really starts to show you why we believe we're going to continue to drive record performance going forward. I'll start with our vision, which hasn't changed. It's been the same vision since I joined the company, put in place. There's really four key words that I want to focus on. The statement is, to be the global leader in analog mixed-signal platforms and algorithms that enable architectural and performance differentiation. It's a big statement, a lot of big words in there.
Let me really focus in on a couple of things. One is analog mixed-signal. That's what we do. We don't do the digital. We don't do much system on chip. We really focus on analog differentiation. We also focus on algorithms, and this is really a key differentiating point for us, is that when we look at how we are approaching our customers and how we're approaching systems, we look at both analog and we look at algorithm differentiation. That enables architectural shifts in our customers. We're able to go to our customers and say, "Why are we doing it this way? Why are you designing it this way? If you were able to change something in your system." By changing the analog or changing the algorithm, you can look at it from totally an architectural shifting standpoint.
We'll give you some examples of that throughout the day, but that's an important part of our vision and strategy. Obviously, our goal also is to achieve leadership positions in our target product segments and deliver revenue growth exceeding the high-performance analog sector, and attaining the high end of our stated operating model. As we look at the industry, strategically, we like to focus on what we call mega trends. These are trends that we believe will continue for many years. Not just one year, not just three years, but maybe five years, 10 years. We focused on three different areas. One is IoT, the Internet of Things. One is mobility, smartphones, wearables. The third is hyperscale data centers, cloud, big data, those three areas. Why do we focus on those?
It's just because we think that they're going to continue for quite some time, and they fit our portfolio of competence as well. In the IoT world, we believe there'll be billions of sensors. Billions of low-power sensors will be deployed, and that's going to continue to drive this insatiable appetite for more data. We believe within the mobile industry, there'll be billions of mobile devices continue to be sold in the next few years. On the data center side, we think that there's going to be continue to be more data centers deployed, and really, just because more voice, video, and data IoT traffic will continue to drive that need for more data centers and more data traffic. We're focusing on those. As we look at this strategically, we try to figure out, okay, how can we get more products into the IoT space?
How can we get more products into the mobile space? How can we get more products and technology into the data center space? I've talked for some time about balance. I want to just take a step back and show you, this is our portfolio today. We strive for end market and geographical balance. We strive for end market balance. 34% of our business is enterprise computing, 31% is high-end consumer, 23% is industrial, 12% is communication. These segments all behave differently. As you know, high-end consumer can be volatile, can drive lower gross margins, can be a shorter time to market. There's also value in high-end consumers play. They tend to innovate very fast. It drives innovation need in the company. Equally, some of the other segments, like industrial communication, more stable, higher gross margins, longer time to revenue, though.
Sometimes it takes longer to drive the revenue. Automotive is an example of that. We have lots of momentum in automotive. It does take time for the revenue to come in. Having the balance is a good thing for us. Equally, as I mentioned, geographical balance. You look at our regional balance, you have 50% from Asia, 7% from Japan, 18% from Korea, 18% North America, 7% Europe. A large part of the Asia number is also demand that's really for end demand in other regions of the world. We have a fairly good balance. As I mentioned, with all of the issues in the world today, sometimes that balance that you strive for is you only realize the benefit when there's issues in the world.
I think as I speak to my board about what if scenarios, I think we're in a very good shape from the standpoint of having a balanced geographical outlook. We also strive for product diversification. I've seen in the industry when you have a product momentum and you have a strategy where customers disrupt or competitors disrupt a space, it's always good to have a broader portfolio, some diversification. That's what we've strived to achieve, and that's what we have today. Signal Integrity Products is about 44% of the business, protection about 29%, and wireless and sensing about 27%. We have a very nice product base, broad portfolio base there.
You're going to hear today from the three general managers of this business. The takeaway is simply that in our signal integrity business and really in the data center space, we just have really outstanding portfolio of clock data recovery circuits. You'll hear the name ClearEdge, you'll hear the name Tri-Edge. Gary will talk about that. In protection, you'll hear the name Z and Z-Ultra. This is our protection platform, really targeted at mobile, also now increasingly at industrial telecommunications and automotive products markets. Wireless and sensing, while we have a number of product areas in there, we're going to focus on LoRa. LoRa is really an incredible story for us. One that's taken a long time for us to bring to realization. It's a very powerful story and one that we need to spend some time on.
In our strategy, in addition to these market trends we look at, focusing on the mega trends, we focus on diversification, we also focus on disruption. What I mean by that is we will try to go to our customers and specifically create value in these spaces. By creating value, I mean innovating, by innovating, I mean bringing both analog and algorithm differentiation to this marketplace. You'll see specifically in LoRa, you'll see it also in other areas, that we're moving slightly up the value chain. The opportunity for us to move up the value chain exists because of the ability to disrupt. You'll see that in the CDR space that Gary will talk about, powering the data center revolution. You'll see that in the protection space with our Z-Platform, enabling the next generation of mobility.
You'll see that in our LoRa play, which is really we're focusing on the message that we believe is the DNA of IoT. I mentioned SAM expansion. One of the things that was very obvious to me when I first joined Semtech 12 years ago is that the SAM wasn't big enough. We needed to expand our SAM. I believe we have done that. We've done that through acquisition. We've done that through really targeting new markets. Going forward, I believe our SAM expansion is going to grow even faster, you're going to hear that from the general managers. Today, we believe we have a $3 billion opportunity growing to about a $7 billion opportunity, so a doubling of our opportunity. I see good growth from all of our product groups in terms of SAM growth.
The $1 billion business market and the revenue that we are looking at now, it's clear. We see it in our sights. It's clear to me that the current growth drivers, the three that I mentioned to you, our CDR platforms, our Z-Platform, and our LoRa platform, will get us there. What comes beyond that is we have a lot of emerging technologies based on the same kind of portfolio and markets, just adding a little bit more differentiation, adding a little bit more disruption. Some of this you'll see in the demo room. IoT, for example, the LoRa tags, which we have opened up, we've talked about in the marketplace, and microservices, which you're going to hear about for the first time from our wireless and sensing team, is what we believe can drive future growth for us.
On the data center side, you'll hear about our PAM4 CDRs. On the mobile devices, you'll hear about our Z-Ultra platform and protection. We also did an acquisition of a company called AptoVision. You'll hear about that briefly for Pro AV. The whole area of smart power and energy harvesting. This is all future, it's all emerging, it's all driven from the current growth drivers that we have in place. We pay a lot of attention also to competition. One of the things that I emphasize with the team, the R&D team, is let's focus on what we do well, what we can call a competence, not on what we believe we should be doing or what our customers are asking for necessarily, just what we do really well. A competence for me, it means you do it best in class.
You have the best skills in the industry. As I mentioned, we focused on high-performance analog, which includes circuit innovation, package innovation, driving patents, driving partnerships within that segment. It also means architectural disruption. It means going to customers and figuring out how to make their systems smaller, higher performance, lower power. We will do application-specific products. We're not a catalog company. We don't do catalog products. We do application-specific products, and we'll push the technology envelope. If the envelope says A, B, and C today, we'll try to see how we can do D, E, F tomorrow. We'll push the envelope. As I mentioned, algorithm design. We can create flexible architectures. We bring our applications knowledge to the table for our customers, and we try to create partnerships across our target segments. Briefly, let me touch on LoRa.
We are going to spend some time on LoRa, but we could spend the whole day on LoRa for sure. It is a game changer for IoT. It's a game changer for Semtech in many ways. LoRa is rapidly becoming the de facto standard in what could be a very, very big industry. A few things that I'm going to say about LoRa, and then we're going to, obviously, Marc and Alistair will spend a lot of time talking about the details. We've invested in this technology for quite some time. I would say it's on the verge of about 12, 13 years, actually. This includes the radio, the algorithms, two acquisitions. We've spent a lot of money investing in this space. Today, as we look at LoRa, we see end-node chip sales. We see gateway chip sales. We start to see end-node chip royalties coming from some partners.
We see microservices concept, which today is a concept. We're starting to show it. You can see it in the demo room, and Alistair will touch on this. Very exciting area for us for many reasons. Essentially, what LoRa enables is the connection of battery-operated assets, indoors and outdoors, low-cost communication, low-cost infrastructure, low-cost end nodes, low-power technology, fixed and mobile communications, scalable and secure networks, and public and private networks. There's a lot of stuff in there, and we'll try to clarify what LoRa does, where it's focused, and why we're so excited about it. One of the beauties about it is the SAM is really limited by your own imagination.
We see so many applications across the globe, from smart environment and industrial, protecting against forest fires, protecting against air pollution, protecting against earthquakes, smart metering, measuring electric, water, gas meters, smart city applications for smart parking, street lighting, traffic sensors, advertising displays, security and smart home, smoke detectors, security systems, smart appliances, agriculture, irrigation control, environmental sensing, animal tracking, and then tracking of assets motorbikes, cars, bicycles, kids, pets. This isn't just a list that we've made up. We have customers around the globe that are in production today in these areas. The key message is that the SAM is really kind of unlimited. It's early. It's starting to be, what I would say is, starting to emerge and become real. We are really intrigued by and excited by the opportunity that's emerging because it's a new industry that's being created. Very exciting area.
Why invest in Semtech? In addition to, obviously, all of the exciting growth engines we have and the excitement around LoRa, we focus on some fundamentals and the fundamentals of growth as an analog company, multiple growth drivers. I mentioned the balance and the mitigation against risks in the industry, and we're a very financially stable company. You'll see we spend a lot of time focused on managing our balance sheet, managing our P&L, making sure that we're doing all the right things to continue to deliver earnings growth for our shareholders. It's a balanced growth story with huge upside. The message I'll leave you with, as I hand over to Gary, is that the journey is really just beginning. As I said, we've never been better poised to take advantage of the secular growth opportunities.
We have world-class products, world-class infrastructure, world-class team, LoRa really is a game changer for us, and I urge you to spend some time, look at the demos, and talk to the team here about not only LoRa, but the other elements of our business, but really understand LoRa as we go. It really is just the beginning. The format of today is that each presenter will present, and then we'll have five minutes of quick Q&A, and then at the end of the day, we'll have all the executives up, and then we'll do a Q&A. If you don't think of your questions now, please write them down, and we can answer them at the end. Let me open it up for questions now for anyone. Great.
Craig, could you use the mic?
Yeah. Craig Ellis at B. Riley FBR. Thanks for the information that we'll see today. I wanted to go back to the Semtech timeline and look at the deal history of the company. Can you just reflect on what you've learned as you've done different size deals, different technology-focused deals, as it relates to how you see pursuit of a SAM that's going from $3 billion-$7 billion, and how necessary do you feel M&A would be as you look ahead to pursue that SAM from the technology and product position that you have today? Thank you.
M&A has always been an important part of our toolkit. We've always viewed it as a good vehicle. We generate a lot of cash. We have a strong balance sheet. We need to use that effectively. Up till now, I would say, my message to the board has always been, let's look for SAM expansion, let's look for diversification. Let's look for how we bring competencies in. I'd much rather bring them in through acquisition than try to develop them organically because it just takes a lot longer, and you always think you know more than you actually know, right? I would say today, the difference in our thinking is really driven by LoRa, because LoRa, the opportunity is so huge. We really, really can't acquire companies to help us with this other than moving up the value chain.
As we look at LoRa opportunity, there probably will be acquisitions, but they'll be more, probably smaller, kind of tuck-in acquisition type deals, and more deals that will kind of, I think, bring us more strategic value in driving the LoRa growth. I think once you hear the wireless and sensing story, and maybe you see the demos and things like that, you'll start to get a sense of why I think that way. Yeah, I think M&A will still be part of our toolkit, but I think probably smaller type of deals. Any other questions? Okay, with that, I'm going to hand over to Gary Beauchamp, driver, communicator on our Signal Integrity Products.
Thanks, Mohan. Okay. Sound check. Yeah? All right. Okay. Thank you, and good morning, everyone. My name is Gary Beauchamp. For a quick bio, I'm an Executive Vice President and the General Manager of the Signal Integrity Products Group here at Semtech. I've been with Semtech for the past six years and with Gennum, a company that Semtech acquired, for 12 years before that, so 18 years total in this area. I have 23 years experience as a general manager and three patents in communications. I founded this particular group back in 2001. I've been running it for the past 17 years, so I have a unique perspective on long-term growth. Okay. For some context, the Signal Integrity Products Group last year delivered $262 million in sales. That constituted, as Mohan mentioned, 44% of Semtech's overall sales. We shipped over 300 million units last year.
Our team is very innovative on the technical side. Mohan discussed that. We talk about that every day. From a marketing point of view, we think we have a really strong world-class team. In total, there are over 300 people focused in this product development area, spread throughout eight different locations around the world. As you might expect, we have quite a number of patents and patents in application. Let me just pause here so we can run a quick video for you to give you some flavor of the Signal Integrity Products Group, and then I'll continue with the presentation.
Semtech Signal Integrity Products Group delivers world-class solutions for point-to-point transfer of ultra-high speed data to the optical communication, broadcast television, and Pro AV markets. The digital world of social media, video streaming, cloud computing, IoT, 4G, and 5G wireless networks have driven explosive growth, resulting in fiber optic connections to the home, global mobile wireless infrastructure expansion, and the construction of hyperscale data centers encompassing several city blocks, housing hundreds of thousands of interconnected servers using 100 gigabits per second fiber optic connections today, and 200 and 400 gigabits per second in the near future. The Semtech optical portfolio of more than 450 products, including ClearEdge clock and data recovery, and FiberEdge physical media devices, enable low-power data transport in applications from less than 1 meter to more than 100 kilometers for the 200 and 400 gigabit per second PAM4 data centers of the future.
Semtech's new portfolio of clock and data recovery products, the Tri-Edge platform, targets the highest volume short reach applications inside the data center. Semtech is powering the data center revolution. As an active contributor to networking standards development, and with more than 300 million optical ICs shipped in 2017, Semtech is a leader in data center, fiber to the home, and wireless infrastructure markets. Over the last 25 years, Semtech video broadcast products have become an integral part of broadcast television workflows, enabling the acquisition and distribution of video content around the world. This portfolio of products expanded with the acquisition of AptoVision and the BlueRiver Technology, leading the way as the Pro AV market transitions from matrix switching to video over Ethernet.
Semtech's differentiated BlueRiver technology offers flexible software-defined AV with up to 4K resolution, transmitting over low-cost IP networks without compression or latency, critical to enhance viewer experiences in public venues such as airports, retail, healthcare, and sporting arenas. Semtech Signal Integrity Products Group offers a world-class portfolio necessary to build the high-speed interfaces of today and meet and exceed market needs of tomorrow.
Go ahead with the video. There are five key takeaways that I'd like to share with you here today. The first is, we are the world's number one IC supplier to the datacom industry for optical transceivers. Second, last year, fiscal 2018, the Signal Integrity Products Group delivered record sales, $262 million. All three of our largest product lines, the CDR, PON, and PMD, they also all delivered record sales. It wasn't an isolated area driving our growth. It was fairly broad. In the current year, fiscal 2019, we're forecasting this to be the ninth straight record year of growth, driven by the 100G data centers. I'll expand on that in the presentation. I'd also like to share with you that in all of our market areas, the data rates are increasing. They're increasing significantly.
Why that is important is as the data rates go up, the need for signal integrity content goes up, we are seeing that in our sales. Lastly, we have mentioned here a few times the acquisition of a company last year, called AptoVision. This particular acquisition has significantly increased our SAM by about a third. It is a very meaningful acquisition to us. It puts us in a particular space where we have an absolutely unique and disruptive technology. That is very important, and I will be spending some time on this particular opportunity that we have in front of us. I am going to first start off by providing some introductory slides to get you a little familiar with the Signal Integrity Products Group, and I will follow that up by speaking to four of our six market areas. The agenda is shown here on the slide.
First, though, what are Signal Integrity Products? Essentially, they are ICs that help transport data from point A to point B in some form of network and do so reliably. There are three key elements that drive the need for signal integrity. One is data rate, the other is the transmission distance, and the third is the media type. By that, I mean, are you transmitting over fiber optic cable or copper cable or some kind of copper trace on a circuit board? For us, what we are seeing is the key driver is data rate. Shown here on the slide, in four of our market areas, are the key data rates as they are today, and where they are transitioning to or are shortly going to transition to. In the data center area, the key data rate today, the high-speed data rate is 100 gigabits per second.
That is driving our sales to record levels. The data rate there too is moving to 200G and 400G. As data rates go up, the need for more signal integrity content goes up. PON today, the vast majority of PON is at 2.5 gigabits per second. That is transitioning. In fact, has started to transition. For the past couple of years, we have been shipping parts in 10G PON, and that is growing. That is moving to 10G PON. That is the next growth area. Wireless base stations, today, the signals, the front-haul and backhaul communication signals are at 10 gigabits per second. That is moving to 25 and 100 gigabits per second. You have all heard of fifth-generation wireless or 5G wireless. That is going to require higher data rates. In our video broadcast space, the main data rate, today, is three gigabits per second. Through the broadcasting system, that delivers high-definition TV.
12G has started already. That is to deliver 4K or ultra-high-definition TV. You can see the data rates are not increasing by 10% or 20% or 30%. They are increasing by 100%, 200%, 300%. It is very significant and has a very meaningful impact to the amount of signal integrity content that is needed in the industry. I indicated previously, we are in six different markets. They are shown here. There are two main areas. One is data comm, and the other is video. I will be speaking to four of these six markets. I would like to point out that, Mohan made this comment at the company level, the company is quite diversified. The Signal Integrity Products Group is also well-diversified. This has taken quite a number of years to arrive at this kind of state, we do have a very diversified business.
We have diversified products, diversified markets, and a very diversified customer base. We believe this diversification is key to having a stable business, it has allowed us to grow our sales for the past eight years in a row, even in times when you have market or customer, even product supplier issues. We have no customer that's a 10% customer. Last year, fiscal 2018, we had 36 different customers that purchased more than $1 million of our product. Shown in the pie chart, if you look in the top left green area, the pie chart is designed to explain how our sales were distributed by the size of customer in the prior year. In green there, we show that 19% of our sales last year came from customers that were smaller than the $1 million customer mark.
In the light blue on the left, you'll see that 19% of our sales last year came from customers that were between $1 million and $5 million in size. In yellow, we're showing that 15% of our sales last year came from customers that were fairly large, between $5 million and $10 million customers. Off to the right in dark blue, you see that 47% of our sales last year came from the very largest customers. Those are customers that purchased more than $10 million of our products last year. These customers, many of them, especially the larger ones, they're not just purchasing one or two products in a specific area. Some of these customers also play in virtually all of the market spaces that we participate in. We're able to do very large bundled sales with them because we have over 450 products.
We're able to bundle large groups of products that span multiple markets, that gives us a competitive advantage, the bundled pricing in the marketplace. Our market trends and growth drivers, two of our markets are considered the main contributors to our current growth and our near-term future growth. The data centers are contributing today to our growth and certainly well into the foreseeable future that should continue. Also the acquisition of Pro AV is the other growth driver that is just beginning. The macro trends I think you're all familiar with, certainly in the data center space, whether it's cloud computing, social networking, video streaming, and now artificial intelligence and certainly IoT, that's all contributing to the growth in the data center space.
In Pro AV, as I'll explain later, it's the proliferation of video screens that you see all around you now when you go out into public spaces, whether it's in train stations, airports, sports bars, and things like that. Behind the scenes is where Pro AV is being installed to stitch together all the information that's being presented to you on the various screens and to switch it, to change it, to update it, et cetera. That's Pro AV, that's growing quite rapidly. I'll spend some time speaking about that. Our market SAM, in total, it's set to double. Last year, the market SAM was estimated at $460 million. We had over 50% share last year in all of the markets that we play in on average. Our SAM is expected to grow to almost $900 million by fiscal 2022.
As I mentioned, the main growth drivers of the SAM, I'm referring to, are the data center space and Pro AV. You can see the impact of the Pro AV acquisition if you look at the bars in purple. The Pro AV adds a significant amount of SAM, and it's actually a special kind of SAM to us because we have a unique product that's highly disruptive, and we think we're going to make some waves in this particular space. In fact, it's already happening, and I'll talk about that shortly. To sum all this up by way of introduction. This is our data comm performance over the past decade. We've grown from roughly $26 million in fiscal 2009 to almost $220 million last year. That's almost 10 times growth in almost 10 years.
This kind of decade-long growth, it doesn't happen by chance. It's the result of being in the right markets and choosing the right strategy in that market. It's good planning and execution, as Mohan has indicated. Semtech certainly has that. You have to have excellent products. Clearly, you have to do a lot of things well to have this kind of sustained long-term performance. Crucially, you also need to have the key supporting infrastructure that has the actual capability to deliver high-quality products and do so in an extremely dependable manner. That's what our customers tell us every day, that Semtech, as a company, is known as a high-quality supplier that can be relied on to deliver. It's a key ingredient in continued long-term success. We, within the group, we have a day-to-day focus on innovation, quality, cost, and customer care.
We look at that every day. The end result is we have become a very dependable partner for our customers. They tell us so. With that introduction, I'll now discuss four of our six markets, starting with the data center. This chart shows the growth expected in unit volume for the optical transceivers at various data rates. Today, the high volume data rate is at 100G. That's the dark blue bar. This has been responsible for our recent record sales. If you look at the dark blue bar in calendar year 2017, you see the size of the bar. That drove us to record sales last year, record CDR sales. You can see how that's expected to grow over the years. This is from the most recent LightCounting report. Our 100G space, we have ample room to grow.
On top of that, though, there are what's known as 25G data centers about to start and 400G. When it comes to 25G, that's shown in red on this chart. In China, later this year, the 25G data centers are expected to start to be deployed. These are being deployed by companies which are the size of Alibaba, Tencent, and Baidu. They're focused on 25G data centers. That satisfies their needs. That's real near-term growth. We already have products. We're already sampling, et cetera. On top of that near-term growth, there's longer-term growth, which is 400G. That's shown in green. The 400G market is perhaps one or two years away. When you look at 100G, that's here and now, and it's growing. We have some near-term growth, which is 25G on top of that, and then 400G for some longer-term growth.
We're pretty confident that this space is going to continue to grow. Of course it is. The Internet just continues to expand. Today, the 100G optical transceivers are generally some variation of four channels of 25G, one example is shown here on the chart. These are being deployed in the millions per year for data center use today. They've become very cost-effective, very power-effective, and extremely reliable. They've set a really tough bar to beat in terms of the metrics that matter. Our ClearEdge technology, that's our CDR technology, it's become the number one selling 100G CDR solution for these kinds of products. Some of them include integrated TIAs or integrated laser drivers. Others are standalone. We have quite a family of ClearEdge CDRs. Last year, we shipped about 12 million of these 100G CDRs.
That's up from 5.5 million the year before. That gives you an idea as to the growth. If I just back up, if you look at the 12 million CDRs, again, look at that calendar year 2017 dark blue bar, 12 million CDRs in there. There's a lot of room for growth for this market going ahead when you look at the size of the bars going forward. I'd like to comment on the past three years of optical module pricing. The 100G optical transceivers, they've experienced a dramatic price reduction over the past three years. Going from about $30 per gigabit, that's $3,000 per module, to $500 per gigabit, which is $5 per gigabit, which is about $500 per module. We're showing just one line here. It's actually a composite of the three more popular 100G transceiver types. There are several different flavors.
Shown here is the average pricing of the PSM4, LR4, and CWDM4 types of modules. We put it all together on a single line here. The conventional solution that I've shown previously, the 4x25, all of these are 4x25G solutions in some capacity. Our customers have been very clever, and they found ways of getting their cost down and being able to ride this cost curve down. What's happened as a result is what was thought to be a relatively short-term solution, four channels of 25G making up 100G, that's become a longer-term solution. Its success has created this virtuous cycle. The success is breeding success. Initially, there were hundreds of thousands of parts deployed, then millions.
Now it's going to be tens of millions of this type of solution, and the end customers have gotten quite comfortable with the reliability data that they're getting back by having so many units deployed in the field. Success has bred success, we expect this to continue for quite some time into the future, this type of solution. However, the future, we're going to see a change in the data center. The next generation of data center will use a new communications protocol called PAM4. That essentially doubles capacity over the current solutions, which are called NRZ. Three years ago, it wasn't thought possible that you could use an analog CDR for high-speed PAM4 communications. There have been advances in the optics, advances in the electronics, and certainly in many of the techniques, and that have changed that view.
Today, Semtech, we are focused on developing a whole line of PAM4 CDRs called Tri-Edge for 100G, 200G, and 400G high-volume applications inside the data center. This is the lowest cost, the lowest power, and the lowest latency solution possible. Our focus is on high volume. What that means is the shorter reach solutions. By short reach, I mean from a few tens of meters to a couple of kilometers. That's enough to cover virtually all the applications inside the data center. We're not focused on data center to data center links in the future. We're not focused on that today either. Our focus is on the high volume. The linking of the data centers is obviously lower volume solutions.
This is our focus today, and we're in development of many of these products, and we had one of them on display in our booth at OFC a few months ago, which was, for us, that was the star of the show. Turning now to wireless. The quarterly growth in mobile data traffic has been extraordinary, and it's shown here on this chart from Ericsson. If you look at the last four quarters, you see a growth of some 64% in data traffic usage required just over the past four quarters. Ericsson publishes a good report. It comes out once a year, updated quarterly, where they put these kinds of actual statistics in the report, and it's quite remarkable the amount of growth.
This demand, though, is actually outstripping the capacity of the systems, and the industry really is in need of a transformational technology to deal with it. That's what 5G wireless is all about. 5G wireless isn't an incremental change from 4G wireless. It's a completely different architecture requiring a lot more nodes and higher data rates. These higher data rates are going to be 25 gig and 100 gigabits per second. Today, it's at 10 gigabits per second. I'm talking about the backhaul and fronthaul links. What that means to us is today, there are no CDRs required in the wireless fronthaul and backhaul solution. Tomorrow, with 5G wireless, all of the links are going to need CDRs. Suddenly we're moving from a market where we're selling TIAs and laser drivers. In the future, it's going to be TIAs, laser drivers, and now CDRs.
That's effectively going to double Semtech content in the wireless space. That's a meaningful development for us. Obviously, we're very pleased to see that. We have 70% share of the market today, which is today, as you know, it's known as fourth generation or 4G wireless or LTE. We have over 70% share of the ICs shipping in that market today, specifically TIAs and laser drivers. As I've just mentioned, in 5G, we're also going to add CDRs to that mix. We don't have those today. The LightCounting forecast shown on this graph shows that the number of 5G wireless nodes required is quite significant. We're already sampling 5G wireless parts today to our customers. Those in China are being very proactive, and Japan and Korea are also becoming very proactive in the 5G wireless space. This is a technology that will be adopted.
It holds great promise for extremely high speed, very fast, low latency kind of communication for all of us who use smartphones. Turning now to the PON market. The PON market continues to be a good market for us. Last year, we had a record year. This year, there's an increase in the number of units that are being deployed, and that's driven by a major replacement cycle of smart ONUs. An ONU is the endpoint box that resides at the customer premise. In China, they are now replacing ONUs with what are called smart ONUs, and there's a replacement cycle, and that's why you see a bump if you look between calendar year 2017, 2018, and 2019 on this chart. If you look in the dark green, you'll see a bump in units, 15 million-20 million additional units because of this smart replacement cycle.
The smart ONUs will allow the telecom operators like China Mobile and China Telecom to offer new services to their customers. The things like being able to control your smart home remotely from your smartphone or tablet, things like that. Our customers were telling us late last year, in the fall of last year, to expect increased units at 2.5 gig in 2018 and 2019 as a result of this smart ONU replacement cycle, and we're starting to see that now. The light green bars on this chart, that's 10 gig PON, and that has already started. On top of the 2.5 gig smart ONU replacement cycle is the emergence of 10 gig PON. We've been shipping for the past few years, 10 gig PON, in smaller numbers until last year and the year before, we got into the millions of dollars of sales.
That's shown in the middle chart at the bottom. The Chinese government is strongly behind 10 gig PON. They want to increase the bandwidth for their citizens to the home by a factor of 5. They're making a conscious effort to push in that direction, just like they are in 5G wireless. The higher data rate, we're moving now from 2.5 gig PON to 10 gig PON, means CDRs are going to be required in some of the parts. In fact, we already have parts and we're already selling parts in production. That includes CDRs for 10 gig. 10 gig is a sweet spot for us. For the past decade, we've been the runaway leader in 10 gig CDR sales, some 90% share. Also in some of the TIAs and laser drivers.
We understand 10 gig solutions very well, and we're just leveraging our know-how in that space to deliver a product in 10 gig PON. In 10 gig PON, we have about 70% market share, just like we do in 2.5 gig PON. We're strongly positioned in PON, and obviously we're pleased to see the data rates moving higher even in PON. Now I'll switch to our video segment. In July of 2017, we acquired a Montreal-based startup called AptoVision with about 35 staff. They had just begun to deliver revenue, so we acquired them at the right time. They're in a really exciting video space called professional audio video or Pro AV for short. The Pro AV market is large. It's a few times larger actually than our existing video broadcast market. We played in the video broadcast market some 20 years.
Pro AV is quite a bit larger video broadcast, and AptoVision is exclusively focused on this Pro AV market. They have a technology, which I'll describe shortly, which is disruptive, it's unique, and there are high barriers to entry. They have a significant lead over any competition. The existing market uses what we like to call dinosaur technology. It's old technology. It's just ripe for a complete overhaul using modern technology and enter AptoVision. That's what they're doing, and that's why we acquired them. They have something very unique. What is the technology, the AptoVision technology? It's called BlueRiver. Today, professional audio video systems use customized or custom switching and dedicated signal routing systems. You'd have your Ethernet network in your building, you'd have to build another network with custom-built switches to handle your audio video.
AptoVision converts all the audio and video to 10 gig Ethernet solutions to IP. These 10 gig Ethernet switches, they're made in the millions and tens of millions around the world today. AptoVision has developed a technology to convert everything to an Ethernet solution, which is widely available and low cost. The technology has won a number of accolades in the industry, and today it's being either implemented or tested by virtually all of the major players in Pro AV. In fact, just a couple of weeks ago, there was the largest conference called InfoComm was held in Las Vegas, the largest conference for professional audio video. It was a very large success for us and for this type of technology. We have reached the tipping point, as I'll comment on here shortly.
One of the writers produced the following commentary after the conference was over because it was sort of what you'd call the hit of the show. The writer from rAVe Publications says, "No one can argue that sending AV, audio video, over an IP network isn't the future of AV distribution. Of course, it's not going to happen overnight or even in one year, but 2018 will forever be known as the year when everyone in AV was introduced to the AV over IP infrastructure." What we've seen is customers, what we saw at the InfoComm show were some of the customers that had previously heard of this technology, decided for whatever reason not to implement it, coming back to us to say they're hearing that everyone else is implementing it, and they don't want to miss the party.
That's kind of the definition of tipping point, and that's where we're at. We have some very unique capabilities there. Why is that happening now? Why is the disruption happening now? Three things are required. You need the bandwidth to be able to carry the data. You need improved cost over whatever you're doing today. Of course, you need the technology to put the data on a network. When it comes to bandwidth, for the first time, the bandwidth from an IP network, in this case, 10 gig Ethernet, is more than what is actually needed to transfer the high video rate of the day. Today's high video rate is 4K. 10 gig Ethernet is no problem with transmitting 4K. I'm talking uncompressed crystal clear transmission.
The cost, for the first time today, the cost per port of switching with an IP network solution, 10 Gig in this case, is lower than the custom switches that are being built specifically for this market. You need the technology to put the audio and the video onto 10 Gig Ethernet systems. That is what the AptoVision BlueRiver Technology is. It allows the signals to get on and off the 10 Gig IP. It allows even for a greenfield installation. If you install the 10 Gig Ethernet network just for your employees to use Ethernet, to use the internet, you can also use that network to carry all your audio/video needs, your Pro AV needs. Today, what you have to do is you have to install an Ethernet network for your internet. Then you have to install another network for your Pro AV.
There are a lot of advantages here. The future is very clear. This is where the future lies. You will have seen it everywhere, this Pro AV. Airports, train stations, sports bars is a big application today, in stadiums of all different types, in operating rooms. Hope you do not see them there, but they are there. Classrooms. In fact, in our demo, you will see a very good demo of how they are being used in hospitals today, linking operating rooms, et cetera. Classrooms, I do not know if any of you have been to a university lately, but you will see they have all got audio and video plugged in. Malls, industrial control. Everywhere where you see multiple TV panels displaying information, whether it is advertising or useful information for train times and things like that. Behind the scenes, that has all been coordinated and switched. The information is refreshed, it is updated, it is changed.
That is what Pro AV is all about. It is virtually everywhere. In addition to acquiring this, what I like to call it, a gem of a company, they are starting to make sales. They are at the tipping point. We are seeing that now. The market that they are actually in is actually expanding. We are not just in a static market trying to take share, which is our goal. The overall SAM itself is expanding. The reason is, some of the drivers are listed here, the printed signage is being replaced by digital TVs. An example that I think most of us have seen, if you have been into a McDonald's restaurant anytime in the past several years, you will notice that the menu and the pricing is all displayed now on video screens. They used to be printed signs. With the video solution, they can change the breakfast menu to a dinner menu.
They can change the prices. They can change what is being displayed, photographs of the food items, and so on. Behind the scenes, that is what Pro AV is all about. In the past, you could not do that. You can think of a lot of applications where printed signage, just billboards and things, are being replaced by these kinds of video screens. The video screens themselves, they are going down in cost, and they are going up in quality, and that is driving additional adoption. Mainly, what is really driving it is video itself is becoming a very important means of communication and collaboration in our society. In fact, today, traffic over the internet, over 80% of all traffic on the internet is video. It is becoming increasingly popular. Our objective here is complete disruption. We want to displace these incumbent old legacy technology solutions with modern Ethernet.
It's what's happened in other industries. If you look at the way you used to watch cable TV versus Netflix being delivered over IP systems and so on. It's not the first time, but this is the first time in an established market, a professional audio/video market. We're pretty excited about it. We think this is going to be a $100 million product line for us in the next five years, and that's rivaling the size of our CDR space, our CDR market. I should mention, four years ago, our CDR market was my, in Signal Integrity Products Group, my fourth largest product line. Today, it's my largest product line. That's the impact of the 100 Gig data centers, has really moved our sales significantly higher. I should point out here, Mohan talked about specific reasons for making acquisitions and so on, and synergies.
In this particular case, this was a company that they have tremendous algorithm capability. The way they would deliver their product was through an FPGA. They take all their algorithms, and they would put them in FPGA form. Semtech, though, we can turn and are now turning their algorithms, et cetera, into a fixed chip. That means we're driving the cost now, and what that does, it's opening up many applications we couldn't have addressed if a custom chip wasn't developed because FPGAs are much more expensive. There's a really good synergy right there. You're taking a software company, basically, which is what AptoVision are, combine it with a chip-making company, and you put the two together, and you get a lower-cost product, which actually opens up the SAM. We're targeting any number.
We've got a very large customer that's interested in the ASIC solution because they could end up using this technology inside your home in big volume. The only way to do that is to get the cost down. That was a really nice example of synergy between the two. Let me just summarize here on a couple of charts. We see continued growth ahead for the Signal Integrity Products Group. The growth drivers are shown here, just to summarize what I talked about. The continued Ethernet expansion, 25 Gig, 100 Gig. There's 200 Gig, 400 Gig coming. Increasing data rates, that's a big part of our content increasing. In some of our core markets, or our foundational markets, we have strong China backing, 10 Gig PON, 5G wireless, and I just talked about the tipping point in professional audio/video space.
Lastly, we think we're well-positioned for the future. We are powering the data center revolution, and we are disrupting the Pro AV space. We feel really positive about where we're at today and where we're headed. Thank you for listening, and I'll take any questions that you might have now.
Great. Thanks for the overview. Woo Jin Ho from Bloomberg Intelligence. It seems as if China wants to ramp up their optical manufacturing capabilities over the next several years. Given your market share as well as your exposure to China, how do you foresee your market share going forward, given what China wants to do?
Yeah. China is, for quite a number of years actually, they have a number of smaller startups within China have produced devices, chips to try and compete. We need to stay ahead, and we are staying ahead on the basis of innovation. We have to continue to do that. What you're talking about is really a microcosm of what's been happening between the East and West for the last 20-plus years, where the West is trying to stay ahead in any number of industries based on innovation. That's what we have to continue to do, is continue to innovate, to bring out more advanced products at better price points. We've seen the threat from the local Chinese chip suppliers. It's coming. It's not there yet, but it's coming, and we're continuing to innovate and staying ahead.
They're coming out now with products that are about a generation behind where we're at.
Thank you.
I've got a question.
Rick.
Can you talk about the relative cost per gigabit at 4 by 25 data center CDR versus true single lambda 100G? Do those cost points, do they have to get close to parity for single lambda 100?
Yeah
Is bandwidth a consideration and all that bandwidth thing? Thanks.
No. Cost is king. Today, the 4 by 25 gig solutions have become surprisingly cost competitive, and that put a lot of pressure on the 100 gig single lambda solutions. Today, I gave some average pricing for optical modules of 500. If you want to choose one of the more popular ones now, which is one of the lower cost ones, is called CWDM4. That's about a module today, an aggressive pricing is around $300. That's about $3 per gigabit. The DSP solutions will have to compete with that, but they'll have to compete with something else. You can take, I talked about PAM4 doubling the capacity in systems.
You could take one of those 4 by 25 gig optical modules, if you replaced what our ClearEdge CDRs and drop in our Tri-Edge CDRs, which basically run at twice the speed, you could use the same optics. Just pop out the CDRs and replace with the Tri-Edge CDRs, suddenly you've got a module for roughly the same price that's doing twice the data rate. Suddenly your $3 per gigabit or $300 module can come down very quickly in price. That puts even further pressure on the single lambda kind of solutions. There isn't yet a battle because the single lambda solutions aren't there yet from a cost point of view. Certainly from a reliability point of view, they're brand new, the DSP solutions. It'll take some incentive to transition to the single lambda solutions in the near term anyway.
The cost is the biggest incentive, those DSP solutions aren't there yet. A member over here.
PAM4 question. Most of the players that are talking about PAM4 are combining a DSP in their sales solution with a physical layer part. Are you also going to do a DSP or just a physical layer, and how do you see that setting up the competitive state of play if you can bundle both in the sale?
Right. What we are seeing, as well as other companies that we've been talking to, and you see it in the industry, is for the shorter reaches You can use the lower cost solution, which is a CDR-based solution. By shorter reaches, what I mean is, say, up to two kilometers. Most of the links inside a data center can be handled with up to two kilometers. There's very high volume at even shorter reaches using active optical cables, for example, for maybe 30 meters or less. The solutions for shorter reaches, we see it as going to an analog CDR approach. For longer reaches, though, DSP is quite capable. More capable, in fact, but you pay a cost in dollars and power. They're able to link the data centers to take the longer reaches from 10 kilometers up to 80 kilometers, for example, can be satisfied with DSP.
We don't have a crystal ball, but from what we can see is the shorter reaches will be handled with the analog CDR solution, and the longer reaches will be handled with DSP. Our focus is the shorter reach. It always has been.
All right. Thanks, Gary. Folks, let's go ahead and keep our questions to the end. We want to get back on track with our presenters. Go ahead and write your questions down, and we'll cover them when we come back at the end for our Q&A.
Okay.
Thanks, Gary.
Thanks, everyone. Next up is Mark Costello for protection.
Okay. Thank you, Gary. Good morning, ladies and gentlemen. My name is Mark Costello. I lead the Protection Products business at Semtech. I'm very happy to be here today to give you an update on our progress since the last Analyst Day. We're very excited, not just about the growth in revenue and customer diversification that we achieved, but also about the great opportunities that are opening up for us in the future, and that's why we believe we're entering a golden era of protection. You'll see throughout this presentation that success in the protection business relies on having the highest performance devices. What does that mean? It means the best electrical performance in the smallest form factor, and Semtech's been an acknowledged leader in this field for many years.
In FY 2018, we were responsible for 29% of the company's net revenue, and we shipped over 9 billion units. I want to show a short video giving some more details about Protection Products and explaining why the solutions that we provide meet critical industry needs.
Today, the world is more connected than ever. From smartphones to tablets to smart vehicles, these devices give us constant access to real-time information and live communication. This communication is no longer limited to humans talking to other humans, as increasingly large parts of our planet are blanketed with sensors talking to other sensors. As our world becomes more connected, microelectronic chips are becoming faster and more advanced. As this trend continues, the chips enabling these smaller, thinner, and faster microelectronics must scale to smaller geometries, leading to increased sensitivity to real-world ESD and transient voltage threats. Consumers expect devices to consistently perform flawlessly, unaware of the hidden dangers posed by ESD and other electrical transient events.
The simplest daily activities, such as walking across a carpet or tile floor or working at a desk, can generate more than 8 times the amount of energy required to damage or destroy today's vulnerable integrated circuits. For more than 25 years, Semtech's Protection Products group has been on the forefront of developing cutting-edge circuit protection solutions to safeguard the various buses, ports, connectors, and interfaces found in consumer products and electronic systems. Semtech's proprietary transient voltage suppression technologies lead the industry in protecting high-speed data circuits against ESD, electromagnetic interference, cable discharge, lightning, and power surges. These innovative devices help ensure electronic devices do not fail in the presence of highly destructive overvoltage threats.
Working directly with customers and IC manufacturers, Semtech designs and tests its protection products to safeguard the latest IC platforms, supplying electronics manufacturers with the protection devices they need the moment they are ready to go to market. Semtech products feature advanced low-voltage process technologies to reduce clamping voltage during an ESD event, the ability to absorb extreme power surges to protect even the most sensitive ICs, ultra-low capacitance to ensure optimal signal quality on high-speed interfaces such as multi-gigabit Ethernet, USB 3.1, and HDMI, and innovative package designs to simplify PCB layout. Semtech continues to design market-leading protection solutions in astonishingly smaller form factors, like the popular Z-Pak technology, which packages maximum surge handling in the smallest footprints achievable. These robust feature sets and product differentiation have resulted in the shipment of more than 9 billion units per year while providing an impeccable track record of on-time delivery.
Customers choose Semtech's proven platforms because they provide leading-edge protection for virtually every high-performance interface on the market today. Semtech provides solutions to customers on time with industry-leading technical support, ensuring that our customers' latest designs are always protected.
We only employ good-looking people at Semtech. The three key takeaways. Protection is, again, a growth business. This is being driven by the rapid adoption of multiple high-performance interfaces and diversification of both our customers, our end markets, and our applications. We've a proven ability to introduce disruptive technologies, and that's significantly expanding our SAM. All of that translates into strong financial performance in terms of revenue growth and profitability. Just to touch on vision and strategy, our vision is unchanged. It was, is, and always will be to be the technology leader in high-performance protection. Our growth strategy is three-pronged. We grow by solving the toughest protection challenges on the newest interfaces in the industry. We leverage our technology and applications leadership to deliver first-to-market solutions for our customers. Finally, we're efficient and effective.
We maximize return on investment by leveraging technology platform developments across all end markets and applications. Let's look briefly at the dynamics of the protection market. This is a really challenging and exciting time for the industry, and that's been driven by the simultaneous rapid adoption of multiple high-speed interfaces. Whether that's Ethernet at speeds of 10 gigabits per second, USB 3.1, which also has 10 gigabits per second speed, but with Type-C, you now have the ability to deliver up to 100 watts of power to charge your smartphone, your tablet, or your notebook. HDMI and Thunderbolt, which are even faster, up to 40 gigabits per second. All of these transitions are happening as we speak, and this is creating an inflection point in the number of protected lanes.
We see growth accelerating in the coming years up to a SAM of above $1.5 billion by fiscal 2023. I want to highlight three key growth drivers. We just discussed the high-speed and high-power interfaces. That's creating a growth vector that's well matched by the advanced protection capabilities that Semtech has. A second growth driver is silicon process nodes. There's an acceleration of a rate of introduction of new process geometries. Between 2010 and 2015, there was only two really significant introductions. Between 2015 and 2020, there's going to have been four or more. Why is this important? Every time a new process geometry is added to the industry, that's adding a slice of SAM at the absolute bleeding edge of protection.
The third key growth driver is the dramatic growth of industrial and automotive applications, including IoT, which you're going to hear a lot more about today. Put together, this gives us a significant expansion of our SAM. Let's look at the high-speed interfaces in some more detail. At these very high data rates, providing a robust protection solution whilst maintaining high signal integrity is a capability that few companies besides Semtech possess. Effective solutions require advanced technologies that don't just provide low capacitance for signal integrity, but also provide true protection. When we deliver the highest performance and, in some cases like 10 Gigabit Ethernet, the only solution, we can extract a premium from the products we deliver, and this moves us away from the commodity market. We showed this slide a couple of years ago, but it's worth revisiting.
Customers introduce new process nodes to achieve things like higher speed, lower power consumption, and that's great for the end product. It comes with significant risk. As you advance in process geometries, it's less practical to include any significant amount of on-chip protection. That means the latest and greatest products are much more susceptible to damage by ESD or hostile power events. To put it bluntly, it just makes it impractical to have any level of system protection on a chip. You require robust off-chip solutions. Again, every new node is an additional slice of SAM at the absolute bleeding edge of protection. We have a track record of having proprietary technologies available that ensure we have solutions that anticipate future customer problems, and that makes our customers capable of securing their product launch schedules.
To make it a bit less abstract, this is showing an ESD event on a leading-edge process node. We're not talking about an inconvenience. You don't just reboot or reset. This is damage and product returns that will degrade the customer experience and degrade your company's quality image. Robust solutions to prevent this from happening are absolutely critical. We've looked at the growth drivers. Let's briefly look at the end markets. We've had almost two decades of leadership in providing consumer protection solutions. Several years ago, we intensified our efforts to reduce our dependency on smartphones. We've taken our technology base and expanded into adjacent areas of security and surveillance and also smart homes and home automation.
We've made significant investment in time and effort to grow our industrial and our automotive business, both in terms of revenue, but a really significant expansion in our customer base. Finally, IoT, while it's great for Semtech through LoRa, it's great for all of Semtech because these burgeoning applications, sensors and nodes are expected to be long lifetime. That requires the provision of high levels of protection so the customers can have a fit-and-forget experience. The takeaway again is this is significant SAM expansion. Consumer is still an important place for us, and the market dynamics are working in our favor. We know the smartphone market, that the TAM is around 1.6 billion phones per year. Semtech still has the largest share at the high end of that market. Content can be in the $0.35-$0.65 range per phone.
A key point is that we now have design positions at six of the top seven smartphone makers, and that's a significant change from a few years ago. The other interesting thing that's happening as technologies progress from the phone into wearables, smart home devices, tablets, and high-end PCs, we also see more and more opportunity. Things such as OLED, Type-C, NFC. We've all found out from initial introductions in smartphone, and this is effectively a doubling of our SAM, and again, a significant reduction in single customer dependence. I hope you take the time to go visit the demo. Unfortunately, we couldn't bring the real thing up to the fifth floor, but we're highlighting some of our progress in the automotive space. Connected car is often referred to as a smartphone on wheels, and why is that?
There's an adoption of multiple advanced technologies on all levels of car these days. Infotainment was a fairly obvious connection to our capabilities in the consumer space, but we're also seeing significant opportunities in advanced driver-assistance systems such as collision avoidance, lane departure, blind spot detection, and in the future, autonomous driving. Why is this? All of these functions need high levels of signal processing, high-speed CPUs, and high-speed communication. Ensuring long-term reliability requires the highest levels of protection. The automotive market drives some of the toughest challenges for protection designs. We've got a growing portfolio of AEC-Q qualified parts, and these have received validation from both new car manufacturers and Tier 1 automotive electronic suppliers. It wouldn't be a Semtech presentation if we didn't mention LoRa every three slides.
Marc will go into this in some more detail, but this is really an important takeaway. If you think about end nodes, these can be in really harsh environments. You may have a light sensor on top of a lamp pole that is potentially vulnerable to lightning, an agricultural sensor embedded in a field, a wearable device on a human or an animal that's getting frequent repetitive ESD events. Low power and long range, they're all important. Long lifetime is a big deal. You want eight, 10, even more years in these harsh environments for electronic devices. Fortunately, we have a solution for almost every application. When we look at the opportunity, there's at least one antenna, and that has to be protected. In some cases, there may be up to eight high-performance ports that need to be protected.
We've developed solutions that are tailored to LoRa, but applicable to the whole IoT market. Again, Marc will go into a lot more detail, but we're really excited by getting our unfair share of the growing IoT business. Let me talk briefly about why we win, and then we'll conclude by looking at the technology platforms and the financial performance. We talk about Z-Platform. What does that mean? It's the embodiment of our best process, package, and design technologies. Is that enough? No. You have to have extremely deep and experienced system expertise. Why does this matter? Protection is essentially a last-minute business. The testing for ESD and other robustness is done just before product launch. When problems come up, there's very, very little time and extremely high pressure from the customers to find a solution.
That requires knowledge to identify the problem, a portfolio of available capable solutions, and the ability to ramp extremely quickly. We've got a strong track record of delivering in these areas, and that really makes us the go-to partner for many leading companies when they need a protection solution. I've been at Semtech for 22 years. I can't quite believe the Z-Platform is coming up to its 10th anniversary, but this has been a game changer for both Semtech and the industry. We've had nearly a decade of innovation in Z-Platform. That's led to the next generation, which we call Z-Ultra. That's significant because this time we've taken approach of having patented solutions, both for ultra-high-speed bus protection for high-speed interfaces, but also for ultra-high power surge capability. We're reaching extremely low levels of capacitance to meet the latest and greatest high-speed data rates.
Previously, we would only ever go smaller. We looked at what was needed. We looked at our capabilities and said the Z-Platform can be expanded in all directions, and this is important for industrial telecom and automotive applications in the future. We've created the capability of putting over 100 amps of surge handling into a package that even only a few years ago, would only have been considered for ESD applications. We designed and delivered hostile power or electrical overstress protection solutions in this package five, six years ago. Six, seven, eight different companies now make the same thing. By using the Z-Ultra technology, we can achieve two and a half times the energy density in a footprint that's 70% smaller. This is a significant achievement. It opens up new markets. Markets that were either unserved or underserved by existing technologies, and it really is disruptive.
We see a SAM expansion of the order of $500 million. If we put all that together, what do we get? We get strong financial performance on all fronts. Since the last Analyst Day, we've delivered on our key objectives. We leveraged the growth drivers that I discussed, and we've significantly expanded our consumer customer base. That led to a CAGR overall of 13% from fiscal 2016 to fiscal 2018. What's more significant is we achieved an actual higher growth rate of 21% in our industrial telecom and automotive markets because they have long, stable, and more profitable revenue cycles. Looking forward, we truly believe our ability to differentiate will continue to drive financial performance. Our first-to-market solutions for high-performance interfaces and consumer give us a solid revenue base. The Z-Ultra platform will enable even steeper expansion into industrial telecom and automotive segment.
We see a CAGR going forward in the 9% to 11% range. That's why we truly believe we're entering a golden era of protection. In summary, protection is again, a growth business being driven by simultaneous rapid adoption of high-performance interfaces, substantial diversification of our customer base, our end markets, and our applications. Our ability to introduce disruptive technologies and expand our SAM. All of that translates into strong financial performance. Thank you.
Hi, Quinn Bolton with Needham. Just wanted to ask the IC Interconnect acquisition that you recently announced, how does that enhance your protection products? It looks like you've developed Z-Ultra already internally, so what does IC Interconnect bring you? Thank you.
I think it's good we can talk about some of these things now that IC Interconnect is part of the Semtech family. It's probably not recognized. A lot of people think that the entire discrete industry moved offshore decades ago. We've shipped over $25 billion Z-Platform parts, 95%+ of those were made in Colorado Springs. It's not about manufacturing. Specifically, what it brings us is the ability to more rapidly introduce technologies like Z-Ultra to production. We don't have to wait and discuss about priorities. We know what we want to do. We've got the capability, we've got the people, we've got everything we need to accelerate, that's the whole purpose. It's IP, and IP. Manufacturing is good. We are very efficient, or we wouldn't continue to do it.
Having a core team that works for us, that's dedicated to going as fast as possible to bring the new technologies to market, that's the key.
Hey, Mark. Craig Ellis, B. Riley FBR. Thanks for the presentation. I wanted to ask about the diversification in the end markets and how go-to market and fulfillment changes as the business transitions from one that I think a lot of us thought of as being very consumer-centric with smartphones and wearables, to one where automotive, industrial, and IoT become a much bigger part. What does that mean for how you engage with the end customer, FAEs, that kind of thing, and then how you fulfill either through distribution, hubbing, et cetera? Thank you.
Full disclosure, I have a bit of a car addiction problem, so I've been really happy. I just came back from Europe a few months ago, and I got to visit the automotive customers for the first time, and I got to really understand. I was suspicious. Why do they need us? They have three European suppliers who've been there for years. Why do they need us to provide protection solutions? We made some investments a few years ago. This stuff just didn't happen magically. We hired a dedicated marketing engineer. We hired one of the best application engineers in the industry, and we specifically told them, "You see a consumer opportunity, you cannot work on it." We always had the capability to address this market, but other things sucked the oxygen away. This is really the power of focus.
We have the distributors in place. We've taken good people and put them on this to make it grow by giving it the attention it needed and deserves. We're starting to see that. Obviously, it's early days because as you know, the time to revenue is long. Every quarter, we're starting to set record levels, and that should just continue going forward. I think we have everything in place. We've got the infrastructure, we've got the team, we've got the channel. What I learned on my trip around Europe is we've got the ear of the customer.
All right. Thanks, Mark. We're going to go ahead and take a 15-minute break, and then we'll come back with LoRa and CFO's presentation. Everyone join us in the other room where you had breakfast, and then we'll get back here at 10:30 A.M. Thank you.
Mike, test. Yeah. All right, good. All right. Good morning, everyone. After the British accent, Canadian accent, Scottish accent, now let's see how you cope with a French accent. My name is Marc Pégulu . I joined Semtech in 2006, the same time as Mohan, our CEO, and I've been through all the wireless stages of Semtech and the acquisition of LoRa and where we are today. I'm now running the Wireless and Sensing Product Group as Vice President and General Manager. In Wireless and Sensing Product Group, the product development is serving a vision of connected sensors, tags, and actuators. Our products include smart power management, proximity sensing, and LoRa wireless. All those areas are driven by the same growth vector, which is the Internet of Things.
I would like to start the presentation by inviting you to follow a short video presenting wireless and sensing and what is our vision of the Internet of Things.
Semtech wireless and sensing products deliver innovative solutions for the Internet of Things with high-performance, low-power communication for LPWAN connectivity and efficient low-power solutions for proximity sensing. LoRa wireless products developed by Semtech connect billions of sensors, tags, and actuators today in industrial, enterprise, and consumer applications worldwide. LoRa technology is bolstered by the rapidly growing LoRa Alliance. A strong worldwide ecosystem of telecommunication companies, equipment manufacturers, system integrators, sensor manufacturers, and semiconductor companies driving adoption of the LoRaWAN open source communication protocol. Created just over three years ago, the LoRa Alliance now registers over 500 global members. This profound growth has enabled the adoption of LoRa in key regions around the world, positioning LoRa as one of the most commercially viable solutions for LPWAN Internet of Things. Operators in major global markets are implementing LoRaWAN networks to support IoT connectivity of LoRa-based devices.
Alibaba Cloud's adoption of LoRaWAN networks in China is only one example of the rapidly growing global ecosystem of operators.
First of all, I would like to thank Semtech for their support of Alibaba in the LoRa ecosystem. Semtech's cooperation with Alibaba is all-encompassing, whether in the chip field, in the network access field, or in the cooperation with partners. They have given us a lot of help. Alibaba also hopes to work with Semtech to provide LoRa networks to more IoT users and serve the public. Thank you.
Like the DNA of IoT, the ability of LoRa to self-replicate and adapt to the variety of LPWAN Internet of Things use cases is making it the solution for billions of future devices. By offering a solution that is optimized for IoT communications, customers can depend on Semtech to build the connection between remote endpoints, picocells, and gateways, then transmit that information to the cloud, ultimately leading to intelligent business decisions based on real-time data.
It's cool to be a part of the innovation that's pushing that into households and making that a standard protocol that everybody's going to know about.
The beauty of it is that it's both interoperable and competitive.
LoRa has really allowed us to be competitive, offering much lower cost structure for uplink technology than traditional cell networks.
We have solutions for facility management, motion sensors and door sensors in buildings. We also have more industrial stuff like soil moistures on farms, parking sensors, water meter, electric meter. They're just plug and play, up and running in a matter of minutes to solve any use case in the IoT industry.
I also believe what we're seeing today in terms of active deployments is just the tip of the iceberg in terms of market potential.
What we really love about LoRa, we have for the last few years, is that it's here now. We're able to take this value proposition to market today.
Semtech is driving the IoT revolution by enabling a smarter planet of innovative, connected IoT applications.
Still, please. All right, not only good-looking people, but we have good-looking things as well at Semtech. Okay. As you have heard in this video, we emphasize about the tagline as LoRa being the DNA of IoT. Mohan pointed a very good point as well. It's all about the transformation that we are living today, not only with LoRa, but everything we do at Semtech. The transformation is really key. With LoRa, we really transformed the way we were seeing the wireless connectivity and the wireless business. In addition, it also transformed the way customers now are seeing their own business. They transform completely their business model after having testing LoRa, and they figure out, okay, there are different ways of making money and serving the customer much faster than what they used to be.
For all those reasons, one of the key takeaway that I would like you to have from today is that LoRa is the enabler of low-power connected devices. LoRa has been developed by Semtech. It is a highly performant, highly efficient, wireless digital modulation for connectivity. It connects sensors, tags, and actuators. LoRaWAN is an open source protocol which is defining how all those devices using LoRa are connecting. The LoRa Alliance is developed together with the LoRa Alliance, which is a very powerful entity proliferating the adoption of the technology. LoRa actually fills a very important technology gap. This gap is the one of the low bandwidth from short range up to long range communication. This is a gap which was existing that LoRa is filling, especially for the sensors, actuators, and tags connectivity.
As a consequence of that, the statement that we are making is that LoRa is complementing the cellular and the Wi-Fi businesses. Wi-Fi is operating in unlicensed band. It is specifically designed for short range and high bandwidth type of devices. On the other side, you have the cellular technology, which are essentially addressing mission-critical applications, outdoor use cases with high power type of communication. What we do with LoRa is we are serving high volumes, B2B, B2B2C applications in unlicensed band to connect sensors, actuators, and tags with the lowest power possible and the lowest cost. Just like Wi-Fi, LoRa is operating in a license band. Just like cellular, it is a highly secured connectivity, because security is very important in Internet. It's all about the trust that you can have between the data communicating from the end devices up to the application server.
Let's speak a little bit more about what is LoRa. In short, LoRa is very simple. It simply connects devices to the cloud. This is what it does. It does this in a very unique way. First, it's about the range that LoRa technology is bringing. LoRa can connect devices which are a few meters away, up to 20, 30 miles away. We actually even have implementations by customers who have been able to make a connectivity from Earth to space, putting LoRa into satellites. The second part is about low power, and low power is important. To link that to the range, if I make a parallel to voice, if I want to get heard, I can speak louder. That's one way of doing it, but it demands more energy. What LoRa is bringing is we are basically bringing you superpower for your ear.
Your ear becomes super sensitive. You can hear much better. By doing that, it requires less energy. That's really one of the uniqueness of LoRa. We can listen with LoRa at a much, much lower energy. If you compare that to cellular technology, this is about 10 to 100 times lower energy to receive the information. This makes a big difference. The third part is about security. With LoRaWAN, with LoRa Alliance, we have very highly expert companies in the alliance, which have defined the protocol of communication. At the very early stage, we made sure that we have end-to-end security, encrypted messages to make sure that the devices which are sending informations can be trusted. It's about integrity, it's about security of the informations, and the ability for the customer to decrypt the information.
The last one is about the fact that LoRa is also a very simple network topology. This is important because with simple network means that you can deploy the connectivity in many different, diverse ways. You can have carrier-grade type of deployments. You can have picocell indoor type of connectivity, and as I mentioned, even satellite type of connectivity. LoRa is supported really by a unique feature, and I would like just to explain a few of them and why they matter. First, I have said that many times, long range. Why long range is important? Long range is important because with long range, you can significantly reduce the cost of the infrastructure. If you can communicate further, why you don't need so many base stations? You reduce the overall cost of deployment of a network. Power.
Lowest power is very important, because in reality, most of the devices making the IoT volumes are going to be battery operated or even operating without any battery. LoRa is bringing really the lowest power options of connectivity in IoT. Mobility. Mobility means the ability to have a device be received or communicate while it is moving. Most of the devices in IoT are going to move from one place to the other. LoRa is working very well with moving devices. We tested it in having LoRa devices in planes, even in trains, fast trains. What is unique about LoRa is that it can connect those moving devices at a very, very low power energy. Last is about the cost. In order to enable all those billions of devices connected to Internet, you need to have the right cost.
LoRa is a very good option to provide a very low cost of connectivity. Let's speak now about the market. LoRa is designed to address the low power, wide area market of IoT. This is really the focus of what we are doing. It is today, I would say, a relatively small market. It's about 100 million of units every year. It is heavily dominated by energy-related applications like smart meters. You have a few other applications which are increasing, like logistic and building home automations, for instance.
As we see now all the opportunities coming, as we see all the developer community developing solutions and customers coming to us, we believe that very quickly, and in a 5-year horizon, that this market is going to grow significantly and reach 1 billion of units per year with a very diversified of segments from energy, logistic, building, and consumer, and even a lot of new emerging applications. We see so much creativity from the developers, as you will see in the rest of the presentation, some of them are not at all predicted by any analyst. This will bring the size of the LPWAN market even further than anyone or anything that you see right now displayed by analysts, and I will illustrate this a bit later.
In this section, I'm going to show you everything that we have achieved so far and why we believe that LoRa is becoming the de facto standard of LPWAN. There are a few metrics that we are monitoring to see whether we are on the right track. Right now we have achieved the key following metrics. First is about the number of deployed gateways. By the end of fiscal year 2018, which we ended at the end of January, we had 70,000 gateway deployed across the world. Which is opening a capacity of 300 million units of LoRa devices. It's volumes of LoRa devices that could be connected through a LoRa network. The network is now touching 50 countries at the end of in January this year. We had, in January, 50 million units of LoRa devices connected.
Last but not least, we reach more than 500 members joining the LoRa Alliance at the end of January. All of this together with a sales funnel that we are monitoring, that we are logging in our system of $400 million. We believe that what we achieved last year is that we set the foundations to drive the future revenue growth of LoRa for Semtech. Let's go now a little bit more into specifics and by the network, starting by the regional deployment. This is a map that was showing where we were at the end of fiscal year 2017. We had about 20 countries with LoRaWAN deployment. At the end of fiscal year 2018, the 20 became 50 LoRaWAN countries. A significant growth in a little bit more than 12 months.
With all the activity that we see right now on LoRa about operators deploying, we believe that by fiscal year 2020, there will be more than 100 LoRaWAN countries deployed with the LoRaWAN network. The second metric is about the gateway, like I mentioned. We ended last fiscal year with 70,000 gateway. What we see right now and projecting for this year is the 70,000 will grow up to 200,000 gateway deployed. The reason for this is because we continue to see a very good adoption of outdoor microcell gateway. These are carrier grade type of deployment of network. Very interestingly is we introduced a bit more than a year ago, the LoRa picocell, which is a smaller form factor, cheaper, specially designed for indoor connectivity. The volume there and the adoption rate is incredible, actually.
All in all, with both type of deployment, we will reach 200,000 gateways deployed across the world. With gateway comes capacity. The capacity, like I mentioned earlier, is about the number of end devices, LoRa end devices, that could be connected to a LoRa network. At the end of FY 2018, we reach a capacity of 350 million units of LoRa devices. By the end of this fiscal year, this capacity will become 1 billion unit of LoRa devices. Where do we stand today in terms of actual connected devices? We ended FY 2018 with 50 million of units of LoRa devices connected to a LoRa network. By the end of this year, we'll be around 80 million LoRa devices connected to a network.
As we see all the opportunities coming and all that, we believe that we are going to cross the 150, 160 million units of devices by end of FY 2020. It's a very strong momentum, basically, we have been almost doubling the volumes every year, with LoRa, which is very, very good. The next metric is about the ecosystem. I have to say, this is probably one of the best accomplishment that we can witness here. It's about how much the technology is endorsed and how much is transforming the business from other companies. We created, a few years ago, the LoRa Alliance. The LoRa Alliance is becoming a world IoT organization. The reason for that is because it has been really constructed from a business perspective.
The goal was not to just to get the best technology, but it's to make sure that we have the right solution for the market at the right time. The quality of the LoRa Alliance is all about the members. The fact that the members are delivering all the layers of an IoT solution. Those layers are the following. It's again around the DNA. D for in the devices, we have other companies, semiconductors company like Semtech, who are also delivering LoRa endpoint solutions. Microchip is an example. STMicroelectronics is another one. We have large module makers like Murata delivering LoRa solutions, but also sensor providers such as Schneider, such as Bosch. They're all part of the LoRa Alliance, and there are plenty others. The network. Really what is interesting about the network is the diversity of the type of network provider.
We have mobile network provider, like Orange is a very good example of a mobile network operator, who has endorsed LoRa and have now a nationwide coverage in France. More recently, we got cable network operators with the venture of Comcast deploying LoRa. Even more recently now, we have the cloud network provider who are joining the LoRa Alliance. In March, we got Alibaba, who joined the LoRa Alliance, and they are now one of the board member of the LoRa Alliance. Very recently, like a week ago, two weeks ago, Google Cloud joined also the LoRa Alliance, which is a very strong endorsement. We have network equipment provider like Cisco. Equally important is when you have the device, when you have the network, it's all about the application.
The key here to be successful is to make sure that we also close the loop, and we have the system integrators are part of the LoRa Alliance family. This includes company like Wipro, like Tata, Capgemini, or IBM. IBM was here from the very beginning of the alliance and they are now delivering solutions as well. With all of this, we really believe that not only we are the fastest-growing alliance, but the LoRa Alliance is becoming a world IoT organization. As we see all the momentum, we set some milestones for this year, and those milestones are going to make LoRa the LPWAN de facto standard. It is supported by the number of gateway. I mentioned 200,000 gateway, opening a capacity of 1 billion units of LoRa devices, 100 countries, 80 million connected devices, and the strongest LPWAN ecosystem.
After going through this kind of education phase and the metrics, I think what is really important is for me to bring the testimony of where we see the volumes of LoRa adoption today, where we see the emerging applications bringing large volumes, and how do we see the future of LoRa. What is it going to open? What kind of business we are going to disrupt with the technology? Let's start by where we make volumes today. The big volumes in LPWAN are right now about smart meter. This is a very good illustration of why LoRa is adopted. We have right now across the world about 30 million smart meters connected with LoRa. The reason why LoRa is used in those applications is just because it delivers a higher performance, and it enables a lower infrastructure cost. What does it mean?
It means with LoRa, when they are adopting LoRa, they can enable lower cost smart meter. This is very important because there are so many emerging countries who are now implementing this kind of smart meter, and all of them require connectivity. If you lower the barrier to entry for both guys with LoRa, you can deploy much, much faster. LoRa is benefiting from a few features. The deep indoor underground connectivity, because a commonality with all the metering space is that most of the time, the meter is very difficult to access. It can be underground sometime. The second part is about the light network infrastructure. There is no need for repeater. It's one of the main reason why LoRa is adopted, because it's such painful to install repeaters in a network, while all of that disappear with LoRa.
The last one is about the operational cost, because a lot of those meters, especially water and gas meter, are operating on battery. If you have to send people to the meter to replace the battery or the consumer to change the battery, that's just a nightmare. With LoRa, you are basically extending your operation life on the battery by 10 times. In addition to that, the utilities which are deploying a LoRa network actually have a huge asset. The reason is once you have a network, well, there are so many other applications that you can connect onto the same network. It's not just about the metering. There are so many other things that you can do. This is what I call the adjacent segment. With the same network, the smart meter network using LoRa, you can connect other sensors.
This enables you to deliver additional services. Like for instance, it can be about water leak detection. If you have a water meter system network, well, why don't you provide as well a capability of identifying where the water is lost in the water distribution network? This is a huge savings. Like in California, we are seeing about 30% of the water is lost because of those leak. Well, now you have a very simple way to put some sensors and help to detect where the leak is and come and repair it much faster. For gas, it's all about safety. Well, once you have the network, you can install gas leak detection. You can also install valve controller, a cutoff to shut down the gas in case there is a leak. For electricity, it's about load monitoring, the power line surveillance.
You monitor the load to make sure that there is no blackout and all that. All of this becomes possible, and we are really lowering the barrier to entry. The same network can cover multiple applications. Now, I said that I would share as well where we see the new adoption. There are several segments which are emerging very, very quickly, and we've seen that over the last months. One of them is the smart home market. LoRa is adopted in this segment because it is simpler and more effective than other technologies. The use cases in a smart home that LoRa is serving are convenience for the consumer, safety for the consumer, but also the home insurance. As illustrated in this slide, the smart home use cases are both indoor and outdoor.
This is really a uniqueness of LoRa, because it's low power and it can do the long range, while unlike other technology in the home, like Zigbee or Z-Wave, where LoRa is perfectly tuned to serve those two types of applications. If you take the convenience, like light control, home air quality, but also you can control your outdoor irrigation system. When it comes to safety, you can connect motion sensor, you can control your door lock, you can have smoke detector. When it comes to insurance-related applications, you can have mouse trap, you can have water leak detection, but also you can have outdoor termite sensor. I put this in the category of insurance, because it is transforming the business model. Especially in North America, but also in some other countries like Japan, a lot of the construction are made out of wood.
Termite is a big issue. Most of the time, every 5, 7 years, you have to treat your house against termite. This is quite expensive. Now the companies who are actually delivering the chemicals to treat your treatment have invented a new business model. They are coming to the consumer and saying, "Okay, look. You are going to sign an agreement, a subscription with us, and we will guarantee that if anything happens to your house for termites, we are going to repair it for free." How can they make this business proposal? Well, they are installing around the house some sensors capable of detecting the termite before it comes into the house. Before anything happens to the wood structure, they will send a team just to clear the surrounding area of the house. That's why they can make this new business proposal.
It's all about now customer retention, subscription and other. It's a very good example of how a technology can transform a business activity. We want to go further. We want to drive the technology to enable new segment. LoRa tag is a very, very good example, because with LoRa tag, we believe we can enable what we call the disposable segment. We are going to make the technology so much low power and low cost that we will be able to connect devices just for the use of a very short time. It can be a week, it can be a day, or it can be one communication and that's it. Allow me to give you a few examples.
If you take the print media industry, where a lot of people are saying, "Okay, with digital, this is a dying industry." Well, I think we can clearly rejuvenate this industry by putting LoRa tag inside the media industry. The media industry is earning a lot of their money with advertising. Imagine now that you have a tag which enables the advertiser to get a feedback about the engagement. How many customers have liked or have been interested in a special product? By putting a LoRa tag, you are connecting your printed magazine to the cloud, and you get a very important information about the engagement of the customer. It also enables as a consumer to receive much more relevant information, instead of receiving so much advertising in your mailbox, where now you get much more relevant about what you want to receive.
It's one example where we believe we can completely disrupt an industry. The second example is about logistics. Imagine you have a LoRa tag on every parcel, on every letter. Well, with that now, not only you can track your device, but you can also know when it has been opened. When, for instance, you can contact your customer to offer an additional service. All of this, it becomes possible because we can bring the technology so cheap, so easy to access. The last example that we have is about compliance. If you take the drug industry, quite often, it's all about, okay, checking that you have a real product, real drug, that it has not been opened. It's not fake, but authenticity. While with a simple seal, with LoRa tag, all of this becomes possible.
You can trust what is being shipped from one place to another. This is a compliance segment. Along those years, Semtech has gained a lot of experience in this segment. We talk with a lot of customers. We believe actually that it offers Semtech some additional growth opportunities. We are now going to speak in the next session about the new business model that are opening to Semtech. Those business model are focused on the installed LoRa devices. To speak for that, I would like to invite Alistair Fulton, who is the VP of marketing for the LoRa activity. Alistair.
Thanks for seeing us this morning. As Marc said, I lead product management and marketing for our LoRa business unit. I actually joined Semtech six months ago as part of our doubling down on LoRa as a growth area, as Mohan was outlining. I've spent the last 25 years in the connected device space, which sounds a terribly long time. Long before anyone came up with the term Internet of Things, there were strangely things that were connected and sent data. Lots of people like myself building solutions that took that data and did something interesting with it. I've spent the last 10 years in the IoT platform space.
The first eight years of which I spent with Microsoft, where I founded the Azure IoT platform as an incubation, actually, then took it through various cycles of development. Then I spent the last two years building the Lumada platform for Hitachi, which Gartner recently rated as the leading industrial IoT platform, which was quite nice of them. Which kind of begs the question, why go away from that space toward the edge more? I think it's based on a very simple conclusion that I reached some time ago, which is, when you look at the platform space, there are about 500 or so IoT platforms in market. You can debate how many of those are actual real platforms and how many are PowerPoint. The problem of what to do with the data, how to drive analytics around IoT, has largely been solved.
Somewhere within those 500, there are the winners. The problem of how to connect all of the devices to these platforms, where all the data comes from, remains unsolved. I think a lot of what you see and hear today, and demos that you see, will help you understand why we think LoRa is that solution. That offering a simple, low-cost, long-range solution of connecting devices in a simple way, really is the key to unlocking value within the Internet of Things. As you know, Semtech has long provided low-cost GPS-free location services. Today, I really want to share the work that we've been doing to deliver more value to the LoRa ecosystem. First, before I get to that, I'd like to step back a little bit and consider the problem that we're really trying to solve as an industry, and why we think it's so important.
Currently, and from experience, delivering an IoT solution is difficult at best. A solution developer needs to be very much a jack of all trades. He or she has to figure out how to design and develop custom devices. They have to figure out the application layer, what's actually running on the device, how is it controlling the actuators, the sensors, et cetera. They need to figure out how to manage those devices on a network, as Marc said, these devices typically are far away from human touch, either underground or in a remote location. They're expected to run for five, 10, 15, 20 years and more. So figuring out actually how you maintain the life of that device is a very significant business problem for a lot of our customers.
Then finally, you have to figure out how to connect that device to a network in a reliable, intervention-free way, such that it will send the data that it is designed to collect time and time again over the many years that it spends in service. In addition to all of that, as if that wasn't enough, you also need to figure out the often more challenging problem of, okay, how do I actually make money? What's the business model? How do I go to market? How do I sell my product? How do I price my product? That's one of the reasons why you see a lot of startups failing in the IoT space because they have excellent hardware design skills, but actually when it comes to the business side, they fail.
Very often, and in many ways, I think, that level of complexity is what has held up the development of the IoT as a whole. I think the IoT is one of those industries where the volume's coming, the volume's coming, it's only been in the last three or four years where we're really starting to see this opportunity be realized. It's complexity that has held back that process in many cases. LoRa is uniquely positioned to address these challenges, for three primary reasons, in my view. First of all, the ecosystem around LoRa is significantly simpler. As I said, there is a Semtech. Semtech's central role in the ecosystem means that we can deliver end-to-end integration from Silicon all the way up to services much more rapidly than any other technology in the IoT space.
That is a very important factor when it comes to simplifying some of the development challenges that I was talking about. Secondly, the ecosystem around LoRa is much more open than other ecosystems, including the cellular ecosystem. LoRa is, as Marc was saying, by definition, an open platform. It uses unlicensed spectrum. So there are more opportunities for much more flexible pricing models and value-sharing models, pivoting away from a model where the connectivity is the thing that's charged for, which breaks a lot of business cases in IoT, to a model where you're actually charging for the value that an application delivers to customers, maybe even wrapping that into a broader service offering.
Finally, LoRa devices, by virtue of the fact that it is an open protocol, they have a standardized communication protocol that works across all of these devices, the millions currently and the billions in the future. They also have a consistent end device model. The reason that that's so important is that, inherently, developers are trying to pick the easiest path to the solution of their problem. With LoRa, they only have to learn one end device model. That applies across all of the industry sectors that IoT is relevant for today. They also, and this also is very key, they have an extraordinarily large market of end devices to target, without having to relearn a technology, relearn development skills, and essentially retool their businesses. Turning things to the slide.
To date, when you look at what Semtech has been doing, as Mohan mentioned, our track record in the development of LoRa is long, and I think, highly influential. We've previously been working to support LoRa by providing industry-leading silicon. Our IC portfolio is at the core of everything that we do. We've also been providing device reference designs. Let me make this work. That's, again, designed to help the industry over the first hurdle of, well, how do I take the transceiver and the micro? How do I do the basics? We've also been providing end-to-end use case support, so working with the ecosystem to identify their end-to-end use cases where LoRa is really delivering value, helping to promote those. Going forward, we think there's more work that we can do to accelerate the growth of LoRa even further. That falls into three categories.
Firstly, by really starting to drill in and identify which areas of solution development drive timeline and cost. Historically, building an IoT solution from start to finish, it takes anywhere between 18 to 24 months. That's a problem, because it takes a long time to get to return, and of course, upfront, there's quite a significant amount of investment. Part of our role, we see, is identifying those areas which really cause pain today. The second focus is really, how do we abstract some of that complexity from the developer? How do we hide it in the box? How do we simplify so that they don't have to worry about it? Much as with mobile application development. Mobile application developers don't need to worry about what's in the box, and they have a very standardized development interface. In IoT, until now, that's not been the case.
What we think we need to do is provide basic application building blocks that significantly simplify the process of building an application. I'll talk to you about what that means in a moment. Thirdly, the importance of the ecosystem around LoRa is hard to understate. The many companies that have invested significant amounts of money, and focus around this technology, those folks are key to us. As much as what we do, how we do it is important. Our focus is very much on how do we simplify and create opportunity for this broader ecosystem of partners to go out and build incredible things that their customers value. Oops. Oh, that's interesting. Hmm. Oh. What are these areas of friction and what do we mean when we say application building block?
We think there are several areas where Semtech can really add value, and some of you, I think, have already seen at least one of the demos outside. The first area we're very focused on is the little blue dot, location. Where am I? I think we're all familiar with being able to take our mobile phones out of our pocket and look at a little dot that tells us where we are. Knowing where things are is one of the most fundamental aspects of the Internet of Things. In many cases, it is the use case. Actually being able to reliably, cheaply, and accurately plot a position allows customers to manage risk, it allows them to deliver efficiency gains, provide contextual insights, et cetera. Today, building a multimodal location solution is actually rather difficult.
For use cases like this typical supply chain from factory all the way through to home delivery, in this case of a television. There's no one single method that actually provides location through that end-to-end span. Bluetooth and Wi-Fi SSID, they're great solutions for indoor location, provided that you're prepared to either invest in the infrastructure or deal with the end-user intervention that's necessary to provision a device to a Wi-Fi network, deal with it when it falls off, et cetera, which is actually the failing of many in-home IoT solutions. When the dishwasher falls off the network, grandma's not going to go reprovision it. That device is essentially dead until someone intervenes. For outdoor coverage, GPS is a great solution, as is cellular.
They come at a very high cost, though, in the form of hardware costs, in the form of a need for increased power, and of course, the cellular network operator cost that you have to pay to operate the device. In short, there's no single location methodology which meets the needs of every use case. In most cases, you actually have to combine multiple of these approaches to know where your thing is, your TV, your package, your pharmaceuticals, et cetera, on its journey as it goes down the supply chain, which drives significant complexity and cost into the development of your solution. Our approach is really, as I said, you will see it in the demo outside if you've not already seen it, is to take all of this complexity and bundle it up and turn it into a simple cloud service that you call over an API.
To do so in a way that enables a developer to choose the location method which is most suitable for what they're trying to do at that particular moment in time. There's very much a trade-off between the level of accuracy that you need to achieve and the cost in terms of hardware and power. Developers typically want to balance those two things. Knowing where something is, down to a really refined point of accuracy, is important occasionally, as something travels through a supply chain. Generally knowing where it is allows you to build an algorithm that you can then build more accuracy around. For a developer, taking all of that complexity, abstracting it away, and allowing them to focus on the real value of actually building the application that solves their customer use case, is a very significant step forward, we think, in the industry.
aggregating device and network-generated metadata, like Wi-Fi SSID or Bluetooth beacons, that the device can see around it, GPS readings that the device can derive, and LoRa metadata, like time of flight on network and signal strength. Combining all of that data, sending it back over a LoRaWAN network into a single service hosted on the cloud, makes it significantly simpler for developers to focus their available time on what matters, their customer's problem, and less time on integrating all of these different methods Supporting a custom solution that they've built through the lifetime of the device. Going forward, we're also looking at ways in which we can simplify the device itself even further by building more location-specific functionality into the LoRa hardware itself to reduce the need for external GPS modules, Wi-Fi modules, et cetera.
aimed around reducing cost and complexity of building a solution even further than it is today. As important, as I mentioned earlier, is what we're trying to do is really how we're trying to do it. As Marc mentioned, the ecosystem is one of the key reasons that LoRa is winning today as the preeminent LPWAN solution for IoT. Our approach is centered on providing utility microservices, that take a complex problem, in this case, location, and turn it into a simple API call, that provide basic building blocks that solve just enough of the problem to unblock innovation while leaving plenty of room for our ecosystem partners to take that building block and build their own innovation on top, and in fact, their own businesses ultimately on top. Application developers can use these building blocks. They can rapidly build an application.
For example, something like predictive geo-fencing, which is something we've seen recently, where you can predict when a device is going to exit a control zone based on the direction it's traveling, the speed it's traveling, et cetera. The simple API model that kind of underpins all of this, it runs on Microsoft Azure today, but in the future, it will run on other public clouds. That also creates a monetary flow for developers. You build ingredients or you build applications, there is an existing API management structure which allows you to monetize those applications across a broader base of customers than you could reach yourself. It also provides a revenue share model for ingredient building block providers like Semtech and others.
Today, in these building blocks, a range of different algorithms currently provided by Semtech that take the basic inputs of things like time of flight, signal strength, Wi-Fi SSID from any gateway or any device vendor that runs LoRa and turn that data into accurate geolocation coordinates in a matter of milliseconds. It's an early-stage service today. This is an area that we're investing in going forward, but it's still in beta trial. We're working with ecosystem partners. We believe very strongly in this space of the value of learn by doing, learn with the ecosystem to make sure that we're solving the right problems in the right way to do what we're trying to do, which is really drive a rising tide that drives the ecosystem around LoRa even further.
In the future, we believe that microservices like this are going to be a really key enabler for the wider adoption of not only LoRaWAN, but really the realization of the opportunity of the IoT, the billions and billions of devices that are forecast. We think that this is a critical enabler. Beyond geolocation, which is where we started for the very simple reason that, as I said earlier, it is the use case in IoT in many instances, we have a roadmap of other services which address other areas of friction in the development process, as well as creating additional revenue opportunities for Semtech and for our partners. With that, I would like to hand back to Marc to close.
Thank you, Alistair. I would like to conclude this section about LoRa by a few key takeaways. The first one is that the LPWAN market, we see that as the biggest opportunity in IoT. Second one is we strongly believe that LoRa will connect 1 billion devices within a few years. The third one is with what has been presented with geolocations and microservices as presented by Alistair, we are really significantly expanding the business model of Semtech by recurring monetization over the installed LoRa devices. For all these reasons, that's why we are stating LoRa is becoming the DNA of IoT. Thank you. I will take now a few questions if you have.
Hi, Mitch.
Just a quick one on the unit count here. If you're going to do 80 million devices, and that kind of gets you to the $80 million-$100 million revenue range you guys have talked to, it implies that ASPs is basically flat. I guess what would drive ASPs up or down in the future in terms of the number of units?
Right now, the ASP for the device, for the chip we are selling in end devices are between $2 and $1. As we are expanding the technology, we are creating also different version of the LoRa chips. Some will be for high-end devices, some will provide additional features associated with the geolocation, and some will be for much cheaper applications like LoRa tag. We'll have a span, a panel, I would say, of different LoRa devices with different price point. With volumes, of course, the price will, of course, go down.
All right. Thanks, Marc.
Thank you.
Good morning. My name is Emeka Chukwu. I guess I'm up here to end the cycle of foreign accents. I want to take a moment to thank the presenters before me, Mohan, our CEO, Gary Beauchamp, GM for SIP
Mark Costello, the GM for protection, and Marc Pégulu , the GM for wireless and sensing. I really want to thank them. The reason I'm thanking them is because I think they've really made my job very easy. All I have to do now is to come up and show you what the financial performance is going to be like. Before I get into that, a few housekeeping items. As you know, we're making a whole lot of forward-looking statements here, and I'm sure you guys know that the actual results will materially differ from some of these estimates. Please take that into account, and the company undertakes no obligation to update all these estimates. In addition, we will be referring to or we have been referring to a whole lot of non-GAAP financial measures. These non-GAAP measures are not intended to replace the appropriate GAAP measures.
Please be aware of that as you think about the non-GAAP measures that we are reporting. If you do need to have a reconciliation to the GAAP measures, we do have a reconciliation at the end of the presentations, or you can come talk to me at the end of my presentation, and I'll walk you through that. Moving forward, there are a few things that I'd like to leave you with by the time I'm done up here. I think you have heard the stories from both Mohan and the general managers, the strategies, and I would hope that by the end of my presentation, you would have come to the conclusion that we do have a lot of growth drivers that are driving our revenues today.
In addition, that we have some emerging opportunities that will continue to drive our revenues beyond our $1 billion target. We've also talked about having a whole lot of value-added platforms that is bringing a lot of value to our customers. Because of the value that we're bringing to our customers, they're willing to pay us for it. We should expect to see our gross margins being sustained at the current levels. We also expect to see them expanding in the future. More importantly, we expect to see our operating margins expanding significantly. We will continue to generate a lot of cash. That is one of the things that we've done very well, and we expect to continue to do that and hence have a whole lot more opportunities to make sure we continue to grow both organically and inorganically.
Very quickly, let's review what has happened since the last time we were here. We were back here two years ago. Since then, from FY 2016 to FY 2018, or looking ahead to FY 2019, based on the consensus estimates, the expectation is that we would have grown our revenues by 30% in this timeframe. That our gross margins, because of our growth drivers from FY 2016 and looking ahead to the consensus, would have expanded 140 basis points. All in that same timeframe, we've been able to grow our revenues 30%, expand our gross margins, and we've held our operating expenses flat, basically. As I go through my presentation, I will clue you in on how we've been able to do that while continuing to invest in our growth drivers.
More importantly, I think within the same timeframe, our earnings per share on a non-GAAP basis has grown 156%. That is more than five times the rate of the revenue growth. When we talk about having leverage in the model, you can see that we are actually showing that we have leverage in the model. Going forward, we still believe that we're not done, that the road ahead is still very exciting. How do we get there? Obviously, if you want to keep growing, you need to have a whole set of foundation or baseline that is very stable. We have a set of core products that are very good and very stable, and we believe in the next five years should continue to do very well for us. Our powerline and wireless base stations.
Actually, as I was listening to Gary Beauchamp, I was beginning to think that maybe this is a little conservative because it does look like our base station business is going to be a growth business going forward based on 5G deployments. Our broadcast video is a very stable business for us, and we expect it to continue to do that. Our legacy protection products being sold into handheld devices will continue to do very well for us, and we have some power and high-rel products. The key takeaway here is that we do have a solid foundation of core revenues that should allow us to continue to grow our top line. Moving ahead, on top of the other reason that I think we continue to grow is when you look at some of the few attributes of the revenues that we get.
Gary talked about not having any significant customers in his product line. Well, the good news is that across all of Semtech, we do not have any direct customer that drives more than 10% of our revenues today. We do not have any product that drives more than 10% of our revenue. 55% of our revenues are done through our distribution. Definitely, that means that we have a lot of customer engagements. We have a lot of people who have enjoyed doing business with Semtech. That is another reason. Given how broad and diverse our customer base is, given the balance in the end markets, that is another reason why we feel very good about our abilities to continue the growth momentum that we have seen. Let's talk about the growth drivers. We've been talking about this now for a while.
Of course, with the 30% growth that you've seen, there is definitely validity in the fact that these are growth drivers. You've heard the story of the IoT space, our LoRa platform, that is doing wonderful things and should continue to do wonderful things. Gary Beauchamp talked about the data center, the fact that as the speed and the data rates are going up, you need more CDRs. That is something that has driven a lot of growth for us, and we expect that to continue. You've heard from Mark Costello, our Z-Platform, that is really allowing us to expand the market opportunities that we have in the handheld space. We're expecting our growth engines in the next five years to grow at the rate of 20% on a compounded average rate.
These two areas I would believe should get us to $1 billion. We're probably not going to stop there, right? We've talked about the emerging growth areas, and we believe these emerging growth areas will take us beyond the $1 billion mark. Some of these growth areas are the ones that you just heard about from Marc Pégulu , and Alistair. We've talked about a lot of tag opportunities, right? That is no revenue today. We're still working on those, but we're very excited about them. We've talked about a lot of micro services, the geolocation. That is something where we have very small revenues today, but we believe it's going to be a significant opportunity for us.
We've talked about a lot of IP licensing, and this is where we enable other semiconductor companies to also sell LoRa, and in return, we get a little royalty from them. We don't have that much revenue from that today, but it's also an exciting area for us. We talked about a TAM for CDRs driving the 200 gig and 400 gig data centers of the future. That is very exciting for us. Gary spent a few time talking about professional audio video, expanding our design for the company significantly, bringing some disruption to that space. He's talked about $100 million of that business in five years. That is exciting for us. We talked about the Z-Ultra protection platform, expanding the market opportunity for us, expanding our TAM by $500 million. That is very exciting. We didn't talk a whole lot about smart power and energy harvesting.
That is going to be a coming attraction for the next Analyst Day. We have a whole lot more to share about that. With the emerging growth areas and with the growth areas that we already have, I hope that you can see a path like we see to $1 billion and beyond. Our record gross margins has been very stable. We expect to sustain the stability of the gross margins. One of the reasons I'm sure all of you care about gross margins is because it is a key indicator of the value that you bring to your customers. If you're not bringing a whole lot of value to your customers, I'm sure they're not going to pay you for them.
The fact that we've had our gross margins pretty high and expanding means that we continue to bring good products, we continue to bring good value to our customers. Part of the reason we think we're going to sustain this is because our end markets are very balanced. We sell into the high gross margin end markets of industrial. It's going to get better with LoRa. We sell into the high-end market of communications. Our enterprise computing, which is where we include our data center revenue, is pretty good gross margin. Our consumer business is also very good gross margin. It is at the lower end of our 58%-63% target range, but it is also a very decent gross margin, and it's getting better with some of the initiatives that we have with the Z-Ultra platform. We continue to bring out new products.
We know from experience that the ways to sustain your gross margin is to continue to innovate and bring up new products. When you have the guys from China trying to get into your space, you have to move on, and that way you're able to maintain your business and keep your ASPs and your gross margins up. We usually do not talk about operational excellence, because at the end of the day, we don't really believe that allows you to expand your gross margin. Operational excellence does allow you to sustain your gross margins. Being able to drive your cost down even helps you to mitigate the impact of ASP erosions. We've talked about the infrastructure investments that we made a few years ago. It's helping us to ship more volumes without significantly increasing our manufacturing overhead.
Now you know we have gross margins that are at record levels that we expect to sustain, we also believe that we should be able to expand them from the simple fact that a lot of our growth is coming from areas with gross margins that are already at or above the current levels. When you look at the emerging product areas, those are all areas that have high gross margin, the IP licensing area, the micro services, the Pro AV space. All those areas are areas that are above the current range that we have today. We do believe that when it comes to gross margins, we can sustain where we are. More importantly, we expect to have an opportunity to expand our gross margins. Moving on, in terms of operating expenses.
Our operating expenses have been benefiting from the investments that we've made in the past. In the past, we put in new infrastructure tools, we put in new ERP systems, new CRM systems, new Workday systems. On the G&A side of things, we have been able to continue to scale the company without a need for adding additional headcount. We're getting a whole lot of leverage from the infrastructure investments that we made in the past. The other thing that we've been doing on the R&D side, I know one of the few questions that I've received in the past from investors is, are you investing in your business given the fact that your OpEx is flattish? Well, we are investing in the business.
The other thing we've been doing is reassessing the areas that we've invested in the past, and if there are any areas that where we see that the return on investment is not really panning out, we have not been shy about refocusing the resources from those areas to other areas. In the last few years, we divested off our Snowbush IP business. Recently we realigned our power management business and merged it with our wireless and sensing to refocus them away from general power and focus them more on the IoT power, the smart power, where we think and believe that we have the competencies to drive a much higher levels of return from that spending.
We've continued to spend in the growth areas that we need, we've funded a lot of those by shifting resources from areas that are not really driving the right levels of returns that we're expecting into the areas that we need. The other thing that is happening is that Marc Pégulu talked about the LoRa Alliance. Think about this. Semtech is driving a de facto standard for the whole world, and we are doing this without seeing significant increase in OpEx. The reason is because we have the ecosystem that is working for us. We have a lot of partners who are supporting us in trying to make sure that LoRa becomes the de facto standard. A combination of all of that has allowed us to be able to drive our top line, expand our gross margins, but at the same time keep our operating expenses flat.
Going forward, we are modeling that operating expenses are going to grow, they're probably going to grow at about half the rate of revenue growth. What we expect to see from that is a whole lot of leverage, that even though the operating expenses might grow, as a percentage of revenues, they should continue to come down. What does all of this mean? I think the math is very simple. You're growing the top line, you're expanding your margins, your operating expenses in check. There is nothing else that could happen except your operating margins, your operating profits are going to go up. Right? We do expect to see a very nice reflection of our operating margins. If you go back to fiscal 2016, our operating margin on a non-GAAP basis was about 13%.
Going forward in the current year, we're expecting this to be in the high twenties. Right? That is a significant increase in op margin. A lot of that increase is coming from the fact that we are focusing on the right markets. Mohan did talk about the things that are driving top line for us, that we focus our strategies on, right? That is the IoT space, the need for bandwidth, data centers, and mobility. Those three key markets are driving significant top line for us, and we expect that to continue. I've mentioned this before, because we're bringing in a whole new level of disruption to our marketplace, we're providing a significant amount of value to our customers. They're willing to pay us for it. That is a key driver for our high gross margins. Then I have what I call value-based OpEx.
I always tell the GMs, I don't have a problem with OpEx as long as you show me the returns, right? If you're getting the profit and you're showing the returns, you can spend any amount of money you want. Because of all those combinations of high growth, stable to expanding gross margins, and operating expenses that is very well managed, we expect to see a significant increase in operating margin. When it comes to cash flow, a lot of you do know that one of the things that we do very well as a company is grow our cash. It is very simple. Everybody knows what the game plan is, we do execute very well. Of course, our cash flow generation comes from being able to grow the top line and hence grow our profitability because of the things I've talked about before.
I believe we have what I call a best-in-class cash conversion cycle. We collect money very quickly. I think our DSO is probably in the mid-to-high 30s in terms of days. Our inventory target model is 90-100 days, and even though we're slightly ahead of that, I do believe that in the near future here, we're going to be within this target range. We manage our disbursements very well. The management of working capital is a focus. It's something that we pay very good attention to in the company. Our CapEx is now being estimated to run at about 4% of revenue as we go forward.
We had previously been targeting 5% of revenues in terms of CapEx, because of the investments that we've made in the past and because of the top-line growth that we're expecting, we do believe that our 4% of revenues in CapEx should allow us to continue to support the top-line growth that we're projecting. With all the cash that we're generating and that we're going to generate, we look forward to putting more of that into growing our top line, but also returning a little bit more back to the shareholders in terms of our stock buyback. The board of directors recently authorized a $250 million increase to our repurchase program, and we expect to use it. We expect to use it to help offset dilution from employee grants.
Every now and then, opportunistically, depending on what the valuations are doing out there, we may get a little bit aggressive in terms of our buyback. More importantly, though, a lot of our cash is going to continue to go into funding our growth. Mohan and I talked about doing strategic acquisitions. We do believe, like he said, acquisitions is a part of our toolkit. You heard one of the examples, the acquisition of AptoVision. An acquisition that is supposed to bring us a $250 million annual expansion, significant amount of revenues. It was a very small acquisition as acquisitions go. Some people probably refer to that as a token acquisition. That is, it is something that is giving us an additional competence in a video space that we're already in.
One of the things that we're also doing in terms of building out our LoRa business is to support the ecosystem. As you can imagine, with LoRa being a brand-new industry, the IoT space, there's a whole lot of businesses, a whole lot of startups that are emerging to help solve some challenges, eliminate some roadblocks that are continuously coming up. We have strategically made some choices where if we see a startup that we believe is going to be able to help eliminate a roadblock, but they need a little bit of funding, we've made some investments in those companies to help get them going, because the more these guys are successful, the more successful we'll be with our mission of making LoRa the de facto standard.
If we do decide that there is something out there that makes a lot of sense in terms of an acquisition, and it's a larger acquisition where we need to go out and raise capital to get that done, the balance sheet is very strong. We do have some debt on the books, but it's actually very low now, all things considered. We have a leverage ratio now of about 1.8 that we expect to be going down significantly very fast due to the top-line growth and the growth in profitability. Our balance sheet is very strong, and it gives us a whole lot of degrees of freedom in terms of how we choose to grow the company. With all this stuff, with a top-line growth that is three points above industry growth rate, that is our target.
I think in my 12 years of being with the company, for the most part, we have actually grown much faster than that. With gross margins that are still very stable and expanding, 58%-63%, we think we are rapidly approaching, if not at the midpoint of that range. With operating expenses that are nicely under control, we expect it to grow at half the rate of revenue growth. I think it has become obvious that the only thing that can happen for us with operating margin is to increase our target range. We are announcing that our operating margin target range on a non-GAAP basis, we are moving it from the current 28%-32%. We now believe that in the next few years here, our operating range should be more in the 32%-36%.
The midpoint of that, 34%, I think we'll probably get there at an annual revenue run rate of about $850 million-$900 million. It allow us to get to the midpoint of that range. Previously, I've mentioned that the CapEx is now 4% of revenues. That's what we're targeting now instead of 5%. Our free cash flow margin, we are also increasing that to 25%-30% from 20%-25%. With that, I would just like to remind you of the things I wanted you to take away from this presentation, and I hope that I have made those points and the case for that. We do have the growth drivers that should sustain our growth beyond the $1 billion target. Because we're moving up in the value chain for our customers, we're bringing more value to them.
We expect to be able to sustain our gross margins and to expand them as well. Our cash generation engine remains very strong. We expect that to continue and that as a result of having a very healthy balance sheet, we do have a whole lot of degrees of freedom in terms of our future growth. Thank you. With that, I will entertain some questions.
Craig Ellis with B. Riley FBR. Thanks for all the financial information, Emeka. What I wanted to do is ask a more qualitative question just on the assumptions that underlie some of the financial parameters that you set forth. Specifically as it relates back to some of the detailed commentary we heard from the general managers and the team. For example, we heard that Pro AV could add $100 million in sales over 5 years. When we look at the core and growth and emerging growth buildup, how do we think about the degree to which that $100 million is fully embedded in your forecast versus shape for reasonable conservatism? Can you comment on the economic and broader backdrop that you assumed as you look out over a 5-year period with the growth profile of the business? Thank you.
Sure. Let me take the last one first. With regards to economic background and trading policies, of course, we don't know what it's going to be, we assume it's going to be steady state. Of course, if we are getting to a broader war with China and we're not able to sell parts to them and stuff, there's going to be some impact, right? That will be a headwind as you can imagine. With regards to the numbers that we actually put into in driving these top-line projections, I'm not going to get into exactly what we put in there, but you can rest assured that Gary and myself are very much aligned in terms of the expectations for the Pro AV market and for the other product lines as well.
Hamed from BWS. How do you expect this ramp to be achieved? Do you need more acquisitions to get there with what your emerging technologies are, or can you do it with what you have right now?
We believe internally, we strongly believe that what we have right now allows us to get to the numbers that we've projected, right? Like we've said, along the way, there could be some strategic talking type acquisitions to really strengthen some competencies and things like that. In terms of being able to get there, we're not factoring in any significant acquisition to allow us to get to those numbers.
Just a reminder, we'll go ahead and open the floor up to any other questions to any prior presenters. I know there were some questions earlier on Gary's presentation that we maybe missed, feel free to ask any questions of the management team at this point.
Actually, let me invite the management team to come up here and open up the questions for anyone on the team on any subject.
Guys, when you hit the billion-dollar target, how big would you anticipate LoRa is as a percentage of that mix?
When we hit the I think the Wireless and Sensing business will continue to expand our size in terms of the revenue contribution to the company at $1 billion. I would expect them to definitely be somewhere ahead of 30% of revenues and maybe between 30%-40% of revenues, something like that.
I guess one of the key things for us is to try to grow our LoRa revenues very fast based on the adoption rate. We're expecting $80 million-$100 million of revenue this year and that to double every year from there. It really depends on, I think how the other businesses do, Jason, that would be really the way to look at it.
A question for Marc. When it comes to competing LPWAN technologies, it looks like the protocols are starting to shake out between LoRa and NB-IoT. When customers choose a protocol, what are some of the talking points or discussion points out on why they pick one or the other?
That's a very good question. I think it's very important to understand, like I said in my presentation, that NB-IoT plays in the cellular license band. LoRa is playing in unlicensed band. In reality, the customer making the decision, they barely ask for technology. They ask for a solution. The key target for us is to make sure that the system integrator providing the final solution to the end customer, picks up the best technology. In some cases, it will be a mix of cellular and LoRa. It's really about the complementarity. From a customer perspective, in that case, this is a system integrator. We believe that the business model that LoRa is offering is much more diverse
You don't have to do business like you were doing with M2M cellular technology. Can open different types of connectivity. It can be paid connectivity, it can be connectivity embedded into other services, even free connectivity in some cases. It's very important not to mix the two. The system integrators are getting to a point where now they understand. In some cases, LoRa type of technology makes much more sense from a business perspective. In some others, cellular, NB-IoT makes a lot of sense.
I think you've talked in the past about once you hit $100 million or so in LoRa revenues, that you will start to see a mix shift from more silicon product revenues to more licensing model. Maybe if you could put some guide rails on what that looks like, what a licensing model looks like, as services turn up and just a sense of what license, maybe royalty rates will be, just kind of a sense of how that model unfolds.
Rick, let me take that. I think there's different types of licensing. We obviously have the chip licensing and the royalties from that, where we already have agreements in place, we know what those royalty models are and what they look like. The microservices that Alistair talked about, I think it's going to take us some time to establish, really what value our algorithms drive out there. Obviously, we're very bullish about it, but we have to demonstrate the value to our customers and then establish in their value system, how much we can extract from that. We'll see. I think it's still TBD. That's why we have it as an emerging growth driver for us. We are very confident, I think you can see from some of the demos in the room next door, that the application space is very intriguing.
Kind of everybody's talking about it in the industry. Coming back to the question on NB-IoT versus, or cellular versus LPWAN and LoRa. Customers generally don't care about what the technology is. Things just have to work well, easily, low cost, low power, and then they're happy. One of the beauties about LoRa and LPWAN is it makes things easy. It makes things low cost and simple. As Alistair pointed out, that drives value for our customers. Our goal is to simply prove that there's a lot of use cases that really require that type of simplicity, that type of low-cost structure, and that will drive the use of microservices and things like that. To answer your question, Rick, it's a little bit unclear for us, that part of it.
I think we're obviously very bullish about it, and that's why we're investing in it quite heavily.
Hey, back to LoRa here. When you guys gave out the 50 million connected devices for last year, how much of that are you measuring as far as being supply chain inventory? How much of that for this year's guidance are you expecting that that's also supply chain inventory as they build up to roll out these devices?
The 50 million is devices in operation today, so they're connected already to gateways. The 80 million we have forecasted suggests that we're going to ship an additional 30 million devices that will be connected this year. Of those, how much will be for proof of concepts, how much will be for real production? It's really difficult for us to say that.
How are you managing the inventory process as the service providers come online, they build up these inventories? How do you manage that? Do you see lumpiness in there?
We don't see a lot. They build out the gateways, and then use cases drive the need for end devices. In any market when there's a ramp up, there'll be initial production runs, right? Their customers will dictate how much they put in place. I don't think that there's going to be a lot of concern about that. From our channel standpoint, there may be some in our distributors holding devices in anticipation of customers ramping up, but we don't see that as a significant issue.
Just to follow up on LoRa, you'll go from 50 million connected end devices at the end of fiscal 2017 to, sorry, fiscal 2018 to 80 million at the end of this fiscal year and projection of potentially 150 million-160 million in fiscal 2020. How are you looking at the supply chain to be able to supply? I assume that you'll be supplying most of those transceivers. If I just look at the difference, you'll do about 30 million units in calendar 2018, and 80 million potentially in calendar 2019. Do you have the supply chain to handle that level of increase, both in foundry, test and assembly, et cetera?
We do. We talked about operational excellence, and in our protection business, we ship billions of units. We're kind of used to the model of shipping a lot of volume. We have foundries, and we have sub cons, and we have multiple sources, and there's nothing unique about this stuff in terms of the technology. Most of the IP is in the design itself or in the patents, and I think the supply chain should not be an issue for us.
Hi, Tore Svanberg from Stifel. IoT connectivity is very fragmented, and there's a WAN, LAN, and a PAN component. Obviously, you're focused on the WAN component of it. Are you interested in the LAN and the PAN component at all, or is that space perhaps a bit too competitive at this point?
Mark touched on home automation, and you'll see from the tags that we have in our demo area, and we've talked about, they come under the category of LAN and WAN, I think. A lot of the use cases are brand new, and Mark and Alistair touched on this, that we're not out there saying, "Hey, it's going to be this or that." We are bringing the technology, and the LoRa Alliance is with a heavy bias at the moment towards more home automation, I think, and more LAN type of applications. Clearly, the value of the range gives you the wider area network coverage, means that there's probably both application spaces are going to do quite well with LoRa. Yeah, I don't think we are saying that it's only one or the other, Tore.
Thank you.
You spoke about the sales funnel for LoRa. Can you talk about that qualitatively, about how that has looked maybe a year ago, two years ago, and how it is today? Can you also talk about, as you go forward and as more client devices become a part of the LoRa revenue stream, can you talk about how you can forecast that funnel and how accurate that can be as you move forward?
Yeah, sure. Let me take this one. I think a couple of years ago, we said that we were in the $200 million type of identified opportunities. What we have seen over the last two years is about 40%-50% conversion rate. The cycle for the conversion is about 12 months, something like that, 12-18 months, depending. We expect the trajectory to continue and to transform 40%-50% of the $400 million within the next year and a half.
I guess there was a second part to your question.
I guess that was it. As you look forward, that's been how it has acted in the past, as you go forward and you continue to project out the funnel, how do you expect that to change given that LoRa is going to grow, essentially double year-over-year? How do you think about forecasting that funnel and the ability to generate the funnel at a higher rate than 40%-50% that you've seen in the past?
Well, the forecasting the funnel, we're going to continue to see that expand as more use cases come on board. The biggest challenge we have is how do you predict the conversion rate from that funnel to end production deployments, right? Some of our customers will do a proof of concept and then never go to production, right? Other customers will accelerate their proof of concept, come very quickly to production. Depending on the segment, depending on the use case, depending on how important the IoT application is in their region, depending if there's any bottlenecks in that location. That's difficult to predict. Our job really, in many ways, is to make that funnel as big as possible, and that's what we're going to continue to do. Then to try to remove the bottlenecks across the different elements of that channel.
Which is, if you look at one of the reasons why Alistair is here and why we hired him and why we brought him in to drive microservices, is our feeling was that's kind of a bottleneck for some of the industrial use cases. Trying to get geolocation accurate quickly, indoors and outdoors and things like that, enables the opportunity if we bring the right algorithms to play for our ecosystem partners to accelerate their ability to get accurate location of an end node device. We'll keep doing that. Emeka mentioned we're making some minority investments and strategic kind of investments in ecosystem partners, and it's mostly to remove those bottlenecks so we can accelerate that funnel to revenue.
Question on the geolocation services. When we were here a couple of years back, you provided quite a bit of detail on what you're expecting on geolocation services. I was wondering if the economics were, if you could provide an update on the economics. If I'm looking at my notes correctly, you were expecting $0.75 per year, per endpoint, as well as you're hoping for a third of attach rate for devices. Any updated commentary on that would be helpful.
I'll start, and then Emeka and Marc can chime in. One of the things we have recognized over the last couple of years is that what customers need from their geolocation activity is, in some cases, more accuracy and more, not just geolocation outdoors, but also indoors, different types of location. We've kind of moved the model, not only being geolocation outdoors, but also to use microservices to provide location within a building, et cetera. That's what Alistair was talking about with geolocation, which you can actually see in our demo area. In terms of the economics, I don't think much has changed there. We still have to prove that we can deliver accurate geolocation at the right price point. It's still, I think, a TBD for us.
We are quite confident with the new approach we have that we're going to be able to do that. Anything else to add?
No, that's about it.
No, I think that's fully accurate.
Just a question for Gary on the Tri-Edge CDRs. Can you remind us again where you are in terms of tape-out sampling to customers, when we'd expect volume production , and are there any sort of roadblocks you have to overcome for the adoption of analog CDRs for PAM4 versus a DSP-based approach? Thank you.
Sure. Thanks, Mark. Just before I answer your question, I didn't completely answer a prior question on the similar topic. We were talking about 400G and DSP and analog. I wanted to point out that we believe both are going to coexist, DSP solution for longer reaches and analog solutions for shorter reaches. As many of you know, we indicated that we didn't go forward with our MultiPhy acquisition, but I wanted to point out they still are our partner, and this is part of the answer to the previous question. We are still developing laser drivers and TIAs for 100G and 400G for the DSP solution. The only thing that's changed is we're simply not acquiring the company, but we're still working with them.
I wanted to point that out because they do have some really good technology, and we've been developing parts to go along with their technology, some of our PMD parts. The answer to your question on Tri-Edge. We have some of our products that were on demo at OFC, which I believe was in late March, early April of this year, and it drew a lot of customer interest. We've continued to work on, we have a couple programs on the go right now. One is for 400G and one is for 200G parts. We will be sampling those later in this year. Before the end of this year, we'll be able to sample the 200 and 400G solutions, and the 100G will be a derivative of one of those.
We expect to see them go into production in the following year and start to derive some revenue later in the following year. I hope that answers your question. Yeah.
Craig Ellis again with a follow-up for Gary. Gary, in your presentation, you provided a number of examples across all the different use cases where CDR content is rising, and that creates a nice tailwind for the company. What happens with ASPs as that content rises? As we think about the total opportunity, how does the blended impact of units and ASPs impact the revenue that we had expected?
Yeah, we think we're going to stay ahead of that curve. There's always ASP pressure. We're always looking for ways of reducing our costs to keep our margins whole. As volume goes up, especially with the mega data center players, the Google, Facebook, Microsoft, and Amazon, those are the four big players. They put a lot of pressure throughout the industry on delivering well-priced products to the market. We work with that constantly. We see some impacts to our ASPs. It's something where we actually see that in many of our products. We try and keep ahead of that curve, at least try to minimize the erosion on margins by driving our own costs lower. The net is we believe revenues are going to go up even though we're going to see some pricing pressure. The ASPs will come down.
Makes sense. The follow-up is for Mohan. Mohan, I wanted to go back to your comments to kick off the meeting because you talked about the confidence and visibility to $1 billion in revenues and then $1 billion plus. The question is on the point with regards to getting to $1 billion, where do you feel like you've got particularly good visibility? Where is your confidence relatively higher? Then on the $1 billion plus, how big is plus?
$2 billion, Craig.
Okay. That's a good number.
The confidence actually across these three business units is pretty high. It's tough for me to sit back and say LoRa is any different than the others because the confidence level is driven by the ecosystem many ways and the LoRa Alliance members, Google joining the alliance. Google is not going to join an alliance if it doesn't feel that there's some value there. Same with Alibaba and those big companies that we talked about. LoRa is kind of in a special place. I think the data center business and what we are doing with our CDR portfolio and how we've got great customer traction and some of the things that are happening in that space is just very exciting. AptoVision, it's a TBD, but you heard the story and the good feeling about it.
With our protection business, we've been saying this for a while, the market, the customers will come to us when the need is there, and we're starting to feel it now. To me, the confidence level on our protection business is increasing also. I started the day saying that we're in the best place the company's ever been in, and I'll reiterate that point that we are really in the best place the company's ever been in. It's a very exciting place.
Any other questions? All right. Management team invites you to join them next door for a buffet lunch. Also, the demos will be open for any of those of you who missed the demos, either at the break or at breakfast. Please feel free to join us and thank you for your attendance today.