Let's get started. We have a very busy day ahead of us. Welcome to Marvell's 2017 Investor Day. My name is Tom Lagatta. I'm Executive Vice President of Sales and Marketing, Worldwide Sales and Marketing. This is going to be a good day for us. We have over 100 people registered to be in this room. The weather's kind of tempered that a little bit, there's a number of other people out on the Internet. A lot of you people seem to be very interested in what Marvell's doing these days, we've got a very, very informative day for you. I have to go through the safe harbor. You guys have been here before, you know that this is obligatory. This presentation contains forward-looking statements. It involves risks and uncertainties.
These statements are not guarantees of results and should not be considered as an indication of future activity or future performance. Actual events or results may differ materially from those described in this presentation due to a number of risks and uncertainties, including, but not limited to, those detailed in Marvell's SEC filings from time to time. For other factors that could cause Marvell's results to vary from expectations, please see the risk factors identified in Marvell's quarterly report on Form 10-Q for the fiscal quarter ended October 28, 2016, as filed with the SEC on December 6, 2016, and other factors detailed from time to time in Marvell's filings with the SEC. These slides provide financial measures of U.S. GAAP as well as non-GAAP basis.
We believe that the presentations of non-GAAP financial measures provide important supplemental information to management and investors regarding financial and business trends relating to our financial condition and results of operations. A reconciliation in accordance with SEC Regulation G for non-GAAP financial measures used in these slides is available in the financial statement section to this presentation and/or in the investor relations section of our website at www.marvell.com. Now that we proved that a public school education product can read, let's move on to the events of the day. Here's what we're going to do today. Here's the lineup of speakers. Let me take a few minutes to go through this, kind of introduce the speakers to you, then we'll get started. Our CEO, Matt Murphy, is going to kick it off.
He's going to talk about where we're going in the future, the results of all the work that we've done over the past year to get the company focused and aligned on our target markets. Matt spent 22 years at Maxim Integrated, where he joined as a product marketing manager and went through a number of things, including worldwide sales and running all the business units, joined Marvell last July as our President and CEO. Dan Christman is our Executive Vice President of Storage. He will actually talk to you about our storage business. Dan is local boy done good for you guys.
After high school, Dan Christman went to the Marines, fixed helicopters and jets for a while. From there he saw the light, moved to California and joined Marvell and worked alongside Matt Murphy for the next 16 or 17 years. Went off to a startup for a few years because it's the law in Silicon Valley that you have to join a startup if you're in the tech industry. Then he rejoined Matt Murphy in the past year. Matt Murphy brought him back from ESS. Chris Koopmans, he's Executive Vice President of our network and connectivity business. Chris Koopmans was actually a PhD candidate at the University of Illinois. As with a lot of PhD candidates and professors, got into the treasure trove of intellectual property at University of Illinois and started a company called ByteMobile.
Decided it was better to have money than not to have money. Ran his company for a number of years until in 2012 it got bought by Citrix, where Chris Koopmans worked for another three and a half years before he actually joined Marvell in the first half of 2016. Jean Hu is our CFO. Jean Hu, after getting her PhD from Claremont University in economics, worked in the Orange County tech industry. You most recently know Jean Hu, she spent the last 200 years or so as the CFO at QLogic. She actually joined Marvell after QLogic, in fact, was bought by Cavium, and she was made available to us. I'm Tom Lagatta. I am the Vice President of Sales and Marketing. I'll talk about what we're doing to drive revenue growth in our customer relationship. I actually graduated from The Ohio State University in 1980.
Got another engineering degree from USC. I'm absolutely conflicted when it comes to NCAA football. Been in the tech industry for about 36 or 37 years, most recently at Broadcom. Most of you guys know me from Broadcom, where I ran worldwide sales. Before that, I actually ran the networking business. That's the lineup for today. At the end of the day, we'll do about a 30-minute Q&A. Then after that, we'll have a hosted lunch. Without further ado, let's get started. I'd like to introduce Matt Murphy, our President and CEO.
That's great. Thank you, Tom Lagatta. All right. Good morning to all of you. It's great to see you. I know a lot of you had to bear with the weather getting here. It's amazing to see the turnout that we've got. I think as Tom Lagatta mentioned, we had something like 150 people sign up to come live, which is a huge turnout for an analyst day like this. I think it speaks to the interest that's in Marvell. We're excited. I'm excited today to tell you the story of Marvell and where we were and where we're headed. First I want to start off with a question, okay? I want all of you to raise your hand if you've ever attended a Marvell Investor Day before. Raise your hand. I don't see any hands. That's because, guess what?
This is the first ever inaugural 2017 Marvell Investor Day. We're breaking new ground here. I think it's a testament and a sign of how the company has changed, and in particular, how we're thinking about investor outreach, investor transparency. It's one of many positive changes I believe we're introducing to the company. Marvell has been and still is one of the leading high-speed communication SoC companies in the world. It's got an incredible engineering heritage. When you think about the explosive growth that's happening in the cloud and the data center, and the amount of storage that's going to be created and network traffic, the bandwidth, the power, the access that's required to do all that plays directly to Marvell's strengths, and we're very excited about that opportunity, and we're going to talk about that today.
We really provide the essential technology to build the networks of the future. For the next 25 minutes or so, I'm going to give you some of the answers to the most frequently asked questions that I get from investors since I joined the company. From my point of view, it comes down to three things that most of you ask me every time I have a meeting. Here's the three questions. The first one: why did you take this job? Why did you actually go off and do this? Wasn't this a little bit risky? What were you thinking? The second is, okay, you've been in there for a while, three months, six months, eight months, however long it had been at that period of time. What have you found? This company was a black box before. We couldn't get information out of it.
We didn't know what was going on in Marvell. What did you find? You've been in there now. Finally, I think most importantly, this is really where we're going to spend the bulk of my presentation and the day today, is answering the third question. I think that's the question on all of your mind, right? Can Marvell grow? I'm going to answer the first two now, then I'll walk you through some slides. I want to stress to you that again, Marvell is not just a value unlock story. It's not just a restructuring story. It's a growth story. Excited to share that with you today. Let's start with the first question. Why did you do this? Who are you? Where's the company coming from, or where are we heading?
As Tom mentioned, I'd spent 22 years at Maxim, at one company my whole career. It's funny, when I've met with investors, a lot of you think that for some reason that me and the new management team, like we've been doing this for years. Aren't you guys a couple of years in now? I want to remind everybody, as Tom mentioned, I joined on July 11th, 2016. I was fairly early in the hiring of the new management team. Actually, a lot of people who are here today that you'll hear from were actually hired after me. This is a relatively new team. For me, although I never competed directly against Marvell when I was at Maxim, I knew the company well. It's a legendary company in Silicon Valley. It's an iconic company. More importantly, the engineering team at Marvell is legendary.
I always had tremendous respect for the talent in the company, I didn't know much about it. Clearly, I followed the issues that the company was having in 2015 and 2016. When I got contacted about the role, let's face it, I was pretty happy. I was working for a good company. I worked with a great team there. I had a great boss, and I really had no reason to leave. At the same time, as I took a closer look, what an opportunity I thought that Marvell presented, okay? Especially the more I got to know about it. From the outside, you got to imagine, it didn't look great, if you look back to when I started talking to these guys in May of last year. It was a company in trouble. A number of issues, a lot of external issues.
You've all read the headlines, investigations, potential delisting, risk of restatement, declining HDD market, leadership transitions. You can just go on and on and on. I actually wrote down a list of all of my concerns. As I went through the process, I actually ticked off all those risks and concerns and realized what was underneath was a fabulous company, an incredibly capable company. When I looked through, I realized, if we just went in there and got the company refocused, there was a lot of value to be created and an impact to be made. In my career, and a lot of you, some of you know me, actually, from my time at Maxim, some of you don't, but I've had the fortune of always being involved in growth businesses in the company. I joined Maxim, it was about a $100 million company when I left.
We were at $2 plus billion. Along the way, I got involved in a lot of product lines and a lot of businesses, and I know what growth looks like. At the same time, at the end of my career at Maxim, in 2015, I was part of the team that also drove the transformation there. I'm not sure how many of you followed that, but that was actually a similar story where company was focused on probably too many things, spending a little bit more on R&D than it should, and needed to take a closer look at its manufacturing cost structure. That transformation was actually very successful. It got the company refocused, dramatically lowered the cost structure, and improved margins.
A lot of that experience I had going through that, which was very different than being in growth mode, going through transformation mode is a different experience and a skill set. It was highly relevant. Even though I had never been a CEO before, and I'd been at the same company for a long time, I really felt like I had relevant experience. When I looked at Marvell, I thought, "I could do this. I could make an impact." When I talked to the board I basically told them, "Look, I think I'm probably the unconventional candidate, if you look at it from an outside point of view." I was absolutely confident I could make an impact. I've built my whole career around taking on challenges, and this was about the biggest one that I could find.
At the end of it, too, you know what I concluded? That this is exciting, and I thought this could be a lot of fun. Okay. What have I found? I started on July 11th, and I had my own three questions when I got into the company. First was, this company's spending $1 billion on R&D. Where is it going? What are they spending it on? You can imagine there was not really the data systems in place to actually uncover that. That took some work to unpack where all the spending was going, what was it targeted towards, had it been returning or not, was this a good investment? The second was just on COGS.
There's another $1 billion a year being spent on COGS, I always thought that the margin structure of Marvell appeared to be low, relative to the engineering and the IP and the talent in the company. That was another one. Okay, how do we go unpack the COGS? The third question, which I'm going to get to, is, finally, when we get through all this, can we grow this company? When I got in, here's what I found. I think I'd characterize Marvell as the world's, and a lot of this term can get overused sometimes, but I think it's really true, the world's largest startup. The company literally was run like a startup, and it got to $4 billion in sales.
It was an amazing accomplishment that it actually got so large run in the manner it was, but it was highly successful for a long time, then all of a sudden, it wasn't. The first thing was, hey, we got to get a solid management team in place. I'll go through some of the members of the management team in a little bit. But you should know as investors, that's in place. We've got a great CFO, finance is under control, Jean's in place. That's done. We've got a very strong operations head, Andy Micallef, who has very relevant experience from his days at Agere and LSI. Operations, under control. That's done. You'll hear actually from Tom today, we've elected to actually give you a deeper dive into our sales strategy, we've got a very capable sales leader, so that's done.
I could go on and on through the management team, but I'm very confident that we've actually put a very capable team in place quickly to go move this company forward. On the R&D side, when you looked at that spending, I'd characterize it in this way. There was a lot of the spending going towards what I would call large consumer opportunities with very large TAMs and very low margins. Ultimately, when I went through the R&D review of the company, those opportunities didn't have the margin performance, didn't have the returns, didn't have the competitive positioning, ultimately, I didn't think they were going to be successful. I also found a decentralized procurement effort, this $1 billion of spending on things, on stuff, nobody was in charge of it.
There wasn't a coordinated manner to kind of go attack it and get religion around it, so that's something that we put in place. That was the goals, get the arms around the R&D, get the arms around the spending, set a timeline, get the team on board, get the company on board, and go execute. On the R&D side, what we found when we stripped out all the, I'd say, somewhat extraneous spending, was actually a great set of core businesses. Phenomenal set of core businesses. Franchises with strong market share, good margin structure, good competitive positioning. We restructured the company around those, around that focus. It wasn't just a cost cut, it was actually a refocusing on a strategic set of product lines and businesses to get the company back on track.
As a result, and I think you guys have seen it in terms of the announcements we've made, teams made a lot of progress in unlocking value. I think the restructuring that we announced and the execution we've had so far, I'm very pleased with. I think our team's done a great job of doing that, and I think we've gotten credit for that from all of you, so we appreciate that. We've right-sized the R&D, we've got the expenses under control, but I think there's more we can do. As I mentioned, whole team's just still getting assembled and, I think we're still in the early innings here. Let's get to the final question, okay? This is the million or the billion-dollar question is, can Marvell grow? At first, I've got to admit, I walked in and I wasn't sure. I didn't know.
It certainly wasn't obvious from the outside that this was even possible. As I've been in the inner workings of this company now, and I've studied the markets, traveled all around the world, talked to the customers, talked to the engineers in the company, worked closely with the product line people, I'm pleased to report there is a lot to like in this company. I'm confident that we can grow this company organically on our own, and let me tell you why. If you think about the mega trends that are happening, one is there's going to be an explosion in the amount of storage that's created.
I think this is well understood, so I'm not going to go through in the level of detail that some of our customers would go through, because I think they make a more compelling case than we do. Just to give you a frame of reference, back in 2013, IDC came out and did a report on the digital universe and said there was, in that year, in 2013, there was 4.4 zettabytes of storage created. Since then, you go out to 2020, and this is projected to grow something like a factor of 10. Key point here is, look, there was a lot of storage being created. You think about mobility, IoT, industrial IoT, connected car, you name it. This is driving a tremendous increase in the amount of storage, at an exponential rate, doubling something like every 2 years.
You think, well, how does this relate to Marvell? This is a good trend for Marvell. Okay? This is a very good trend for Marvell. For us, whether it's traditional hard disk drives, nearline or capacity drives, new technologies like TDMR, or even on the flash side, traditional NAND, 3D NAND, or other storage class memories. Actually, Marvell is agnostic to the memory. Marvell doesn't make the memory. What Marvell does really well, actually better than anyone else, is we move the data on and off the media faster and at lower power and at higher reliability than anybody else out there. If you think about a throughput issue or a bottleneck in the network, this is actually a real bottleneck.
This is a real problem to be solved because you've got a tremendous amount of storage and data sitting out there, and it's got to come off the media, and it's got to come off the media accurately and at high speeds. This is really what Marvell does. When you look at this trend, and I think everyone can kind of wring their hands about different aspects of the storage market, and is it going to grow or not? How's Marvell going to do? We sort of step back and look at the mega trend and go, there's going to be an enormous increase in the amount of storage out there, and we're well positioned across a number of memory technologies to participate.
The second one is on the data flow, I think this is also a well-understood trend, which is just the amount of data traffic happening. Again, you've got to store the data, and then you've got to move the data across the network. This is the traditional Cisco VNI report. From the consumer to the cloud, billions of devices connecting to the internet, it's creating an explosion of data traffic. Again, through Marvell's networking technology, we become an essential reason why this bandwidth is enabled and can happen. One of the reasons that this data traffic's increasing so much is live video. Actually we just had last week, we had the Snap IPO, marketed themselves as a camera company, which was kind of unusual.
The traditional YouTube, has moved from sort of stored video to their live offering, Twitter Live, and ultimately Facebook Live. Again, what does this have to do with Marvell? You think about Facebook and Facebook Live. I guess Facebook's up to about 1.8 billion users now. They're generating about 8 billion views a day of live video, and these videos are stored forever. Today it's in 720p, it's going to 1080p, and then 4K. How does this all work? Actually, how does this benefit Marvell? How does Facebook Live benefit Marvell? Okay, here's The Rock. I don't know if you guys remember this. The Rock was live streaming via Facebook before the Oscars. He had something like a million views. What's interesting, if you follow the data flow, that data flow goes through the carrier network, right?
That actually goes through Marvell Technology and Marvell products. It goes through our switches, it goes through our PHYs, it goes through our SoCs. That broadcast is sent to a streaming server, then the data is stored in what's called an origin cache. Again, those go through Marvell data center switches and then ultimately into Marvell storage. The subscribers, which is on the right, they'll request to watch the live feed. Could be in your home, which would go through a Marvell home gateway product, which we'll show some examples of that today. Or if you're in your workplace or you're in a coffee shop, you would go through a Marvell enterprise-grade Wi-Fi access point. Ultimately, all this video and all this data is actually stored not once, not twice, but three times. Again, you say, well, okay, great. Get it.
This was the big aha moment for us, when we dug into Marvell, and I think for some of the investors as well, where Marvell's value is, where the IP is, where the margins are, where the core of the company is actually in the center. It's in the center, it's in the cloud, and it's in the infrastructure. The challenge that Marvell had was it was spending a lot of its resources, energy, and focus on the periphery: IoT, mobile phones, TVs, you name it. Consumer-oriented devices sitting on the edge, where didn't have as strong a competitive position, and quite frankly, didn't have the profitability and the franchises.
For us, when we thought about what to do next, we said, look, let's take the assets of the company, let's streamline, let's focus on the center, let's focus on the infrastructure, and try to get this pivoted to build a sustainable, long-term business model out of it that generates healthy margins, cash flow, and all the things you would expect. Now, one challenge was, if you look back, and this was some interesting data. We've started to cut the data little bit differently in Marvell now, and this is just one insight that we got. This actually, we took a look at not just the product lines that Marvell has, but where do the products end up? Is it going into automotive? Is it going into computer? Is it going into mobile phones? We've got a view of that now.
We unpacked just the portion in the past that was going into the cloud and the infrastructure side. Old Marvell, this was about 25% of the company's business. Actually, another big chunk, if you remember, was a big mobile business, which ultimately was shut down, and we have very little revenue left on. This presented a little bit of a challenge. As we've restructured and refocused, and we've seen some nice growth over the last four quarters, actually, this portion of our business now is about a third of our business. We're in a much more healthy place now to grow off of. About a third of our business is in this kind of stuff from an end market point of view.
As you look at our future plans and our growth plans, we're going to give you some proof points on those today, our business model and business plan at Marvell is to actually get this kind of business, these end markets, to represent half of Marvell or even greater. We think this is where the growth opportunity is. This is the view you're used to seeing, okay? This is the cut now by the product lines that we sell. If you look back in the past, storage, networking, and connectivity, those are the three focus businesses of Marvell, and now we've got everything lumped into other. Again, if you look in the past, other was a substantial portion of Marvell. It was not small.
The interesting point was when I got into the company, the R&D investment in that piece of the pie was actually even larger than that. We were over-indexed from a spending point of view and a revenue point of view on to businesses and product lines that did not carry the kind of performance that we were looking for. Again, we restructured, we refocused, we did some de-emphasis, and we did some re-emphasis on the core of Marvell. You can see now, this is getting close to 90% or so of our, actually, our base revenue are in businesses where we have growth plans, we're investing R&D, we're investing sales resources, and we're trying to grow. It's an important point I want to make.
We're not trying to take the company and grow it based on businesses that we're not really in today, that's sort of going to swamp out the businesses that we already have. These are actually businesses that we're in, and we've got healthy investment in those. In the future, we expect that almost all of our revenue is going to come from these three segments based on, one, the de-emphasis of lower margin businesses, and then the emphasis and the investment in some of these growth areas. We think this is an eye-opening story. Just a quick run-through now of Yeah, sorry. There's the change over time. A quick run-through now on our three businesses, and I'm just going to stay at a high level here because Dan and Chris are going to take you through in more detail.
On storage, quite frankly, it's a great business for Marvell. We're the market leader here. We've been at this for 20-plus years. Actually, it was the first business of Marvell, was our preamp technology. Company possesses tremendous system-level expertise, which creates a high barrier to entry for others to enter, and we've still maintained and continue to invest in technology. That's our storage business. Our strategy, and I don't want to steal Dan's thunder, but our strategy is to continue to grow our market share in this business, shift our mix to the cloud, because traditionally, Marvell was more on the client, and we're making progress there. Through some other initiatives in R&D, actually expand our available market in storage. In networking, we've got a long history here as well.
Ethernet PHYs were one of Marvell's original businesses, along with the acquisition of Galileo in 2000, where we got switch technology, again, we've been at this for close to 2 decades. We offer a complete technology platform for networking, which is pretty unique, especially in the enterprise and the SMB area. We'll talk about that and how that plays to our strengths and our growth strategy. Here, really, I was even stunned to kind of see the numbers finally add up. We have a very refreshed portfolio here in networking of new products that are actually getting traction and driving revenue growth. We're going to really leverage that as we think about some of these market transitions that are happening in networking and enable Marvell to be well positioned in the future.
We're also going to talk about, which is I think on a lot of your minds, is what's our play for the data center in networking and how are we going to compete there, especially with some large established competitors. Be clear, the data center is part of Marvell's strategy. I think we've got a thoughtful way to enter that market and participate in that growth as well. Finally, on connectivity, this is one where a lot of investors, I think, have questioned me probably harder than the other two. Okay, you've got this business. Isn't it in a bunch of mobile phones? Why do you still have it? What are you going to do with it? Again, we'll dive a little bit deeper, but remember, we are an established player in this market. We still have substantial revenue.
It was an organically developed business inside Marvell. Again, here we have a complete technology platform of RF, of SoC capability, of firmware and software, and system-level knowledge. We're really a leader when you look at what the current mix of our business is in Wi-Fi is actually what I call enterprise-grade. So it's either for enterprise access points or automotive or really high-end home gateways. So I'd call it the premium segment of Wi-Fi, and we think there's some exciting opportunities there, especially as we've worked our way out of some of these lower-margin opportunities and into a much richer product portfolio going forward. Again, we think by focusing on performance, focusing on markets where it's not just about the standard, but it's about innovating around the standard, we think we can add value.
Again, growth markets like automotive and connected home, we think provide a nice growth trajectory for this business. That's the high level. You're going to hear in more detail from my team about this, but we think we have some nice growth opportunities in all three of these segments. A little bit on the leadership team. Over the last eight months, we've really built an entirely new leadership team in Marvell, and I'd say collectively, we're already operating as a team. These are senior proven leaders in the industry from different companies and different backgrounds. A diverse set of experiences. All of them have operated at scale in the past, either at Fortune 500 companies or large established companies in our business. More importantly, this team has really started to gel. Actually, we're gelling really well.
I think one of the things that sort of brought us all together is we all do share a common vision for the opportunity at Marvell and what this company can achieve. It's actually been a really rewarding part of the journey we've been on to bring in this kind of talent and get everybody working together. As we're seeing progress, it's quite fulfilling. Tom already did a very colorful introduction of several of these people, so I won't try to compete with him. Dan runs our storage business. Mitchell Gaynor, who's here, runs our legal. By the way, many of these individuals are here, so when we have the lunch afterwards, I encourage you to actually interface with not only the presenters, but the rest of the extended management team who's come out to meet with you as well.
Jean is our CFO. Gary just joined us last week from Broadcom. He runs corporate development. Chris is networking and connectivity. Obviously, you've seen Tom up here already, who I've known for over 10 years and was actually the sales VP at Maxim Integrated at the same time he was the sales VP at Broadcom. We go back. Andy Micallef runs our operations and is doing a fabulous job and his team of getting our cost structure under control. Finally, Karen joined us in December as our head of HR. We've got a nice team that's been filled out, and we're very happy about that. I do want to take a moment to acknowledge also in the same time frame and even before really all of this, we also put in place a new and expanded board of directors at Marvell.
That's worth mentioning because the company did suffer from challenges around its corporate governance practices. The fact of the matter is, we've brought in actually a world-class board that's not only new, but provides, I think, great oversight over the company, great support and advice to me and the management team, and have really been champions and cheerleaders for the company as we've gone forward. All of us on the management team appreciate the board's support. Remember, they put in a lot of hours that most of you don't even realize before any of us got here when they got involved with the company and started to turn the ship. I will say, I'll spend just a few seconds on this company is just more than executives, though, okay? It's more than a board of directors.
We have thousands of employees in Marvell, okay? Who, let's face it, can you imagine being an employee of Marvell over the last two years? These guys have gone through a lot of turmoil and a lot of soul searching on what's going to happen. I want to just give them a quick compliment and also let you know that it's been really refreshing to get involved with this company because the employee base actually has been incredibly receptive to change, incredibly enthusiastic about that, and really supportive of all the actions that we've been taking. I couldn't be prouder of how they've done it. This is important to know because everyone's asking, "Well, how did you guys make this right turn so hard?
How have you actually been able to execute and get done what you've needed to get done?" It doesn't happen if you don't have the team behind you and you don't have the crew. I just want to acknowledge also that this was a big team effort, not only for me, the management team, and the board, but the thousands of Marvell employees, many of whom, by the way, are incredibly long-tenured and care a lot about this company. We've made a lot of progress. Back in July, after about six days in the company, I don't know how many of you were on this analyst call, earnings call.
I joined July 11th, and July 17th, I was thrown onto my first earnings call with Rick Hill, where we had to go over our Q1 earnings. It was not a pretty result. They were obviously delayed. We came out and said that we were down to $540 million in revenue. We put out gross margins of 52%. We basically made no money. I think we made like $4 million in that quarter. I think that was a low point. Even when I was coming in, I'm like, "Oh, my gosh. How low is this company going to go?" That was a tough call. By the way, I've now subsequently completed four additional earnings calls. I've actually done five earnings calls in like seven months, which I think has got to be in the Guinness World Records.
If you can find another first-time CEO that did five earnings calls in seven months, I'd like you to tell me who that is. I'm getting experience with this at least in talking to all of you. If you just kind of take a look at where we were in Q1 2017, kind of a little bit of a darker period, you look at the guidance that we just gave in our earnings call last week for our projected Q1 outcome, you can see that actually revenue now is projected at $570 million, which I think was certainly better than consensus, which we were very happy about. I think gross margins, in particular, I think we're most proud of the progress we've made here.
I think if you asked anybody a year ago, "Was Marvell going to be able to guide a 59% gross margin quarter?" I don't think you would've gotten many people that thought that could happen. Again, that's the plan of record. Finally, I think it is a milestone. It's not something that we sort of view as we're done by any stretch of the imagination. I think this sort of mythical 20% operating margin Can Marvell get there and then can Marvell actually go through it is another question, but midpoint of our guidance suggests we'll actually get to a 20% operating margin next quarter. Again, I think the framework of this is, we've made progress. You've all seen the results in the guide. I want to stress, it's happened in a pretty compressed period of time.
I just say we think that there's more we can do. This company has a unique combination of assets. If you think about it's really a leader in mixed-signal design. We leverage it across the entire portfolio. It's kind of the core of Marvell and what differentiates us. The company also is very much a systems company. Its combination of silicon, software, firmware, and actual knowledge of how the end systems are built is a major differentiator for us, and we consider that a major asset. We also have flexible software, and we provide a lot of our own software to customers. You'll hear that from Chris and Dan today, how we do that. We also are a big proponent of open source, and we enable that, or we allow the customers to actually customize. This flexible software model adds stickiness to our products.
Actually, it's a key selling proposition and value proposition when we go to market with the customers. Finally, look, the company was built by a founding team that believed in R&D and believed in innovation. That innovation is alive and well in Marvell. We just, for the fifth year in a row, got this award as one of the top global 100 innovators by Clarivate, which used to be Thomson Reuters. We had an all-employee meeting just this past Monday when I was in California before I flew out to the East Coast. The end of the meeting was we did a panel with the engineering leaders of the company talking about how they're driving innovation, and then we actually had the CEO of Clarivate come up and present the award to the entire team.
Again, you've got this company now that's amassed 9,000 patents and again, has an engineering team and a scale and a culture that I think really lends itself to pushing the envelope in new products and will help us in our growth mode. Our customers compliment us all the time, by the way, about the quality of our team. Speaking of customers, that's another thing I've been really impressed with at this company is despite all the issues that it went through, this is a very loyal customer base to Marvell. They did a great job here as a company on building relationships that are really world-class. I think as the industry consolidates, if you think about it, there are fewer and fewer players, especially ones with the capabilities that Marvell has. They actually want us to win and want us to succeed.
We got some feedback from some of you when we were putting this together about what would be some useful things. A lot of you said, "Hey, it'd be great to hear some customers give a testimonial." We actually reached out to the customer base, and you'll see it a little bit later today. Throughout the presentations, actually, we got a whole bunch of customer testimonials, so you'll actually hear them talk about Marvell and back up some of these points that I'm telling you. You can hear it from them, actually, not me. When you add all this up, it looks like these end markets we're in, they're growing about 6% a year. They're growing about 2x the semiconductor industry, as it turns out.
Semiconductor industry, if you look at most forecasts, is sort of a 3% grower at this point, which is too bad, very different than when I joined the industry 22 years ago when it was 15%-20% growth. Now it's 3%, but the segments that we're targeting are growing significantly faster than the overall semiconductor market. I believe, with the talent of this company, the technology we possess, the position we have in the markets that we are targeting, our plan is to outgrow the market. I would settle for nothing less, and to take share and to get this company back on track and make us incredibly successful. I think if we do that, as investors, you'll be incredibly successful as well. In summary, to answer my last question, can we grow? Look, we have a leading technology foundation in our core businesses.
We typically have a number 1 or number 2 position, and you'll kind of see that. We have strong franchises. We have blue-chip, great customer partnerships, the top OEMs in the world designing us in, pulling for us, and advocating for Marvell. Finally, we've got this set of markets we're now targeting that are aligned to this mega trend of the creation of massive amounts of data that need to be moved and stored across the networks of the future, and we're aligned to the cloud. We think with the combination of all these attributes of Marvell, at least I'm confident that we can grow this company. I'll end by saying I'm incredibly excited to be here, incredibly proud to be part of Marvell, and I'll tell you that at Marvell now, there's a renewed sense of purpose in this company.
A renewed sense of purpose. You walk the hallways of Marvell, you go to our international locations, you can feel it. Our customers see it, and I hope today as you go through the presentations, I hope that you feel it as well. Thank you very much. Okay.
Hello. I'm Dan Christman. I'm the Executive Vice President of Storage at Marvell. When Matt called me last summer and asked me to join him at Marvell, I thought, okay, well, I know Matt really well. I worked with Matt for almost 17 years at Maxim. I knew Matt's character. I knew Matt's skill set. I knew how successful he had been. I knew his character. Also, Matt was a mentor for me for years and years at Maxim. I knew Marvell's technology was great, but didn't know a lot about the company until I talked to Matt and got his insights into what the opportunities were. I also knew myself. I knew my own capabilities and knew my own successes, and I knew I could make an impact.
I can tell you, it ended up being a no-brainer to join, and I look back now and I'm thrilled I made the decision and thrilled I joined Marvell and I'm very happy with the decision. Today, as I stand here and talk about storage, I'm actually very excited to be able to tell you the story of Marvell Storage. It's really a story as old as Marvell itself. Marvell was founded on storage and the read channel technology that's a key part of our business today. As we stand here 22 years after the founding of Marvell, I'm very proud to say that Marvell is still the leader in storage, and we're growing. I think most people here probably know that Marvell is the leader in hard disk drive controllers. I think everyone probably knows that.
Maybe not everyone here knows we're also the leader in merchant silicon for SSD controllers. In the last quarter, we announced that we were actually more than 20% of our revenue and storage now is coming from solid-state drives. This is an area we believe we can continue to grow and actually grow faster than the market. Matt also talked about how data is being created everywhere, lots of data, 10x data from 2013 to 2020, and that's really good for Marvell. More data, you got to store it. Often, you store it more than one time, as Matt showed. We think that we can grow in this market because we think HDD is growing in the cloud. There's double-digit growth in SSD across all the segments that Marvell participates, retail, client, data center, enterprise.
There's also new opportunities that I'll share with you today and some new markets and new technology that Marvell is going to participate in that will actually expand our SAM further. Matt showed you the strategy for growth or the strategy for storage, rather. For HDD, it stands the same. Number one, we're going to increase our share. We're going to increase our share because we have technology leadership that enables us to win more business with our customers. We're doing this by maintaining our business we have today in the client, but we're growing quite fast in the cloud. Data center and enterprise are growth markets for Marvell's HDD. Third, we're going to expand our SAM. I'll talk today a little bit about the preamp business. Preamp is a piece of silicon inside the hard disk drive.
You have at least one preamp that ships at every hard disk drive, and it just happens to be that the SoC, the controller, and the preamp talk a lot together and help make the HDD work. Why does Marvell win in hard disk drive? Number one, we have deep-level system knowledge. Matt talked about this. We've been doing this for more than 20 years. We design for the system, not just for the chip. I'll talk about this in a moment in more detail. We also have technology leadership, where Marvell has the best read channel in the industry. We have the most advanced FIRs. We have the best error correction codes. We also are on the most advanced process nodes, and we'll talk about this more in a few moments as well. The last thing Matt said was our customer relationships.
We have tier 1 customer relationships with all three, actually, in hard disk drives, very close, long-lasting, deep relationships. Rather than me tell you, I am going to let one of them tell you.
I am Dave Mosley, Chief Operating Officer of Seagate Technology. Seagate's a 40-year-old technology company making hard disk drives, the storage solution that the world has needed for all of its digital creations, consumers, all the way to enterprise. For the last 40 years, we have been the leading provider, and we are really proud of this and would continue into the future. Seagate was Marvell's first customer way back in 1995. I was actually in the division where Marvell first rang the doorbell and got their products in the door. Obviously, a deep engineering partnership started then, it is because of Marvell's excellence in technology and what we call the read channel, which is really a communication channel that allows us to process data down to the drive and back off the drive, turning the analog signals on the drive back into digital signals.
Marvell has been our top partner over the many years since then, and we have developed areal density solutions, which allow us to put more terabytes on every platter together for these many years. I think that technology is still very vibrant into the future. With such a long history together, Marvell and Seagate have obviously been working very cooperatively on innovative solutions that have stood the test of time. The most impressive thing that I can think of about the Marvell-Seagate relationship is the fact that it is as ever as vibrant as it ever was. Many suppliers have come and gone in this industry, not just on the read channel side, but all other parts of the silicon have been integrated into one chip, and Marvell's been with us the whole way.
As a matter of fact, I think the silicon solutions into the future will be ever more complex. The interfaces are getting more complex, the demands on the storage products are more complex, and Marvell is actually out there in front understanding that technology, able to leverage it in the future, and a critical partner for Seagate. Data is growing a lot in the digital world, and we all know that. That is against the backdrop of tremendous change in architectures. The PC was one architecture that we all knew really well. Now, as the data locality is changing to the cloud and maybe other applications like gaming or surveillance, the architecture's changed very quickly. You need partners who know the workloads, understand all of the technologies around those architectures, and can help us.
Marvell's a perfect partner in that space, I look forward to a bright future together.
As Dave mentioned there, Marvell's been at this a long time, more than 20 years, okay? We have an understanding of the architectures to be successful in hard disk drive. We understand the technology, and we have technology leadership. If you think about it, a hard disk drive is a very complicated piece of hardware. It's not just a piece of hardware, it's a very complicated system, in fact. Imagine you have a head that's hovering one nanometer over a platter spinning up to 15,000 RPM. It's not just hardware, it's not just software, it's mechanics, it's electronics, it's material science, it's chemistry. This is quite a multidisciplinary piece of equipment. It's pretty amazing if you think that thing works, let alone works reliably and works over a number of years. Marvell, as we said, is a system-level expert.
You see the number of patents we have here, you see the number of units we've shipped. We work closely with our customers to develop new technologies that help them increase their capacity, increase their reliability, lower their cost, improve read/write speeds. As you're going to see later, one other point I'll make on this slide is that this same leadership, the same technology, the same amount of IP, this gets applied directly to our SSD products. It's one of the reasons we've been so successful is the scale and the IP we've generated in hard disk drives gets directly leveraged into our SSD business. We design essentially the brains of the hard disk drive. As Dave mentioned, these are becoming ever more complex. This is not a business you can harvest and be successful. You need to invest.
You need to have scale, you continue to innovate in order to be successful in hard disk drives. Dave mentioned the read channel, the communications channel that basically transfers information back on and off the platter. We have 18 generations of innovation in read channel. It leverages our mixed signal expertise that Matt mentioned earlier. We're also the leader in error correction. Error correction is what allows you to basically put data on the disk reliably and pull it off. As you go to higher areal densities on a disk, you're essentially putting more bits in a smaller area. That means you have a smaller signal, which means it's more prone to be corrupt. Once you put something on a disk, if you put it on there incorrectly, you lose it forever. It's not like in a network where you lose a packet, you send the packet again.
If you don't write it correctly on a disk, you don't get it back. Error correction is extremely important. Also low power. We talk about highest performance per watt. You can also read this as lowest power for the same performance. If you're in a mobile device, lower power means longer battery life. If you are in a nearline or a high-capacity drive, it means you can put more of them in the same rack because you got lower power, less heat. If you're in an enterprise drive, it means you can get better performance for lower power. It also means your drive will be more reliable because there's going to be less heat in the enclosure. Dave talked a little bit about areal density gains and how Marvell helps with areal density gains.
This is an interesting one because when you talk about aerial density and you talk about how you interact with a disk, there's how you write to the disk, and it's how you read from the disk. When you talk about how you write to the disk, one of the things we're proud of is this last year, we started shipping with our first customer, what's called SMR. It's shingled magnetic recording. It's a way to write to the disk. Now, normally, you write one sector at a time. With SMR, the word shingled, basically you think of your roof. It's like you've got one overlapping another. It's a shingle. Here, what you basically do, you write two tracks, one after the other, so two at a time, and the second one overlaps half the track of the first one.
You actually are overwriting data that you wrote already. Through complex algorithms and error correction codes, you don't lose the data. You actually are able now to compress more data in the same area on a drive. What that allowed us to do, our customer at least to do, using their technology and Marvell's technology together, is ship the first mobile 2.5-inch drive, single platter, one terabyte, twice the capacity of the previous generation. We're also working with customers on the next two generations of write technology, what we call HAMR and MAMR. HAMR is heat-assisted magnetic recording. MAMR is microwave-assisted magnetic recording. You're basically adding another element into the drive that can locally either heat from the outside or from the inside the media, which basically lowers the coercivity of the drive and allows you to write into a much smaller area.
Another way of increasing capacity on a drive, we're very far along with customers on these technologies. We believe we'll be first again to ship with those technologies. Last, on the read side, we're working very closely with customers today on what's called TDMR, which is two-dimensional, which actually means you're using multiple heads. I'm sorry, not multiple heads, but multiple readers on a head. Now you can have wider readers. You can use two of them, maybe three of them. You offset them, and you get a bigger signal because you're basically reading two parts of the track. You can basically subtract out the signal, get rid of the adjacent track noise, and have a better system. You'll see that coming.
I can't say when, but we're working very closely with customers on that technology, and we'll be first again on that technology as we move forward. The other thing we're doing is leading in technology nodes. I'll skip the next slide on that. Marvell was first to market with 40 nanometer. Today, we ship majority of our production is in 28 nanometer. Next year, we'll start shipping customers in 16 nanometers. The important part about 16 nanometer is you still get figure of merit increases that are meaningful. Okay, I get a smaller die. What does that mean? Smaller means I can actually either have a smaller die, lower the cost, or I can add more stuff to it. That's really important because, as we mentioned, these are still evolving. You get lower power. We talked about lower power.
It means a lot of things, but it means lower liability as well as longer battery life and lower heat and higher performance. The higher performance metric. We believe our 16 nanometer platform that we'll be rolling out will be, again, industry-leading and will also be more efficient with our engineering, both for ourselves and our customer. Matt also talked about the mix from Marvell moving into the cloud. We're seeing it in hard disk drives as well. It's a growing segment. The biggest piece here is nearline from a growth perspective, which basically is capacity optimized. Imagine multiple disks, multiple platters in a single hard disk drive, maybe six, seven, eight, nine, 10 stacked right on top of each other. These are higher value sockets because it's more difficult to make these work. You need to set up timing between platters. It's very difficult to do.
It's a growing market, Marvell's also growing share in this market. Traditionally, if you look back maybe a year ago, our share was rather low in the cloud. We're very pleased, as we said in Q1, our guidance, that we're very pleased with design wins, and we're seeing ramps in this segment. Out of FY 2018, we expect this to be a much more meaningful portion of our revenue. If you think about this particular market, if you have the drives, the number of drives are growing. They're becoming more complex because you have more and more platters in them. We have less share in their higher ASP. We think all these things line up very well for Marvell in our future in the cloud for hard disk drive.
The other thing I mentioned at the beginning about our strategy here for hard disk drive was expanding the SAM, expanding the content we can have in a drive. There's really probably four major components that are semiconductor components in a hard disk drive. There's the SoC, the controller that we do, I showed you earlier. There's a DRAM that typically sits next to it. There's going to be a motor controller on the drive that helps spin it. The other piece really is the preamp, and this basically sits on a head stack assembly. You can see there's a picture there showing. This is a nearline drive, so you can see there's multiple heads because there's multiple platters sitting in between each of those. In the case of a preamp, you're going to ship at least one with every drive.
As you move into nearline and multiple platters, you end up shipping more than one. More platters means more channels. You need two channels per platter. You need more channels, which eventually means more chips. They're more complex because there's timing issues. You essentially actually create quite a significant amount of additional content that can be shipped per drive. It works very closely with the SoC. The SoC talks to the preamp. The preamp executes, gives the data back to the SoC, then processes the data. They work very closely together. We have a very unique insight into this market if you're a company like Marvell. We actually started shipping our first preamps in 2016. We're seeing tremendous customer pull.
It's a very small part of our revenue today. You can imagine, again, as we take share in a market that's growing with higher ASPs, it's a very important business for Marvell going forward. Matt also mentioned the word agnostic earlier. We know that storage is growing. We know that the amount of data is growing faster than capacity increases can keep up with the data, which means overall, the data market or the storage market is growing. Here we all acknowledge that the hard disk drive, there's some replacement happening in certain areas by SSD. Sure, we see a decrease in the overall market SAM for Marvell there. Keep in mind, this particular graph here doesn't include the preamp, which is about a $350 million market. This is the controller SAM only.
If you were to add in the preamp, it'd maybe be about minus 3% CAGR. The key to take away from here is that overall storage is growing. Marvell is taking share, and we believe we can grow in storage. We'll move to the next one, which is SSD. You see the SSD. This is the growth of SSD. We've broken it out here between kind of client and retail and cloud, which is data and enterprise. This is our fastest-growing business in storage. We stated in Q1 it was more than 20% of our business. Based on the ramps we're seeing, we believe this will be more meaningful coming out of FY 2018 than it is even today. The key on SSD for Marvell is it's a very balanced business for us.
We cover the full portfolio of products in retail, client, enterprise, and data center. That's a big advantage Marvell. I'm going to show you a little bit more about this as we go. Our customers love this about Marvell. By the way, it's not easy to do, especially in the higher end of the data center enterprise. It's very high reliability required there, and it's something unique that Marvell knows because of our HDD experience. Last is we're seeing growth across all the segments. Again, our strategy, as it is in storage at HDD, for SSD, it's the same. Increase our share. We're going to outgrow the market and gain share. Grow in all the segments, not just the cloud, all the segments, although cloud is our fastest-growing segment within SSD. Then expand our SAM.
Like I showed you in hard disk drives with the preamp as a new market, I'm going to talk to you about what we call network storage in SSD. I think Tom later will talk a little bit about automotive, which is essentially a server on wheels, which will also have some storage, but we'll let Tom talk about that. Why do we win in SSD? First of all, we have a full range, as I mentioned, of SSD supply, SSD controllers, and we also have a very flexible business model. Matt talked about our software. He talked about how we're flexible in how we work with our customers. We're flexible on how we make our chips and how we deliver software. I'm going to explain why shortly. We also leverage, I've said this a few times here, we leverage our HDD technology.
We've been doing it for 20 years. A lot of the same technology that's used in hard disk drives is used in solid-state drives. It's a huge leverage for Marvell. Last, strong customer relationships again. I'm going to let one of our customers tell you. This is Nate Steffens, he's the VP of Flash at Kingston, and Kingston is one of our long-term partners in SSD.
My name is Nate Steffens, I'm the VP of Flash for Kingston Technology. Kingston is the largest independent manufacturer of memory, and we also have NAND products, including solid-state drives, USB drives, and flash cards. Kingston's vendors are very important to us as a business, so we like to partner with somebody that's going to be with us for the long term to bring great products to the market, great quality products to the market for a long period of time. We started partnering with Marvell in 2012. We developed some controllers in tandem with them and asked Marvell to develop firmware on our behalf, and they have been great in helping us do so. We buy controllers with them on a regular basis for our highest-selling product, the UV400.
Marvell is very meticulous in their quality, so this particular firmware, developed exclusively by Marvell, allows us to give a great high-quality product in our SSDs. Our SSD business has grown tremendously over the last several years. We are now shipping over a million drives a month, and we feel that Marvell, given the resources that they've allotted to us and given the quality of product that they've delivered, will be a great, tremendous long-term partner with Kingston.
Marvell has developed strong relationships across many of the tier 1 suppliers in this business. We do this because our customers really care about quality, reliability, and performance. They care about our broad product portfolio and our flexible business models. Let me talk to you about that right now. Marvell, as I mentioned, covers all the market segments. We cover retail, client, enterprise, and data center. We're very unique in having this breadth of product and technology in the solid-state drive business. And it's possible really because of the scale of our entire storage business. We also offer a flexible business model. Let me talk about that a little bit.
We have some customers who want to buy a very simple off-the-shelf SoC and full turnkey firmware, which means we provide all the software, all the firmware, the production code, the manufacturing code, the test code, and they can just make a drive and take it to market, right? Marvell does that. We have other customers who want to include some of their own IP in the design, maybe in the design of the IC, maybe in the software, maybe both. Marvell, we do that as well. We have other customers who say, "Here's a specification. I want this product exactly this way, and I want to be the only customer who buys it from you." We do that as well. Marvell's been very flexible with our customers, and this has been very successful for us.
There's a new model that's developing, it's really targeted to hyperscale data center, the AI customers, and this is where the customer will buy or pick or select, if you will, an SSD controller from a vendor like Marvell. We provide with them a software development kit that goes with the chip. They go and they select their NAND, their flash memory from one, maybe two or more suppliers that they choose. They put their own firmware on top of the solution, and then they go pick an OEM provider or OEM supplier to build the drive for them, test it, manufacture it, produce it, and they essentially end up with a custom SSD controller for their data center, meeting their specifications, and in some cases, very flexible.
Marvell is very well-positioned, and we're seeing a lot of traction in this space because number 1, we, as I told you, we have a flexible business model, flexible firmware, flexible software. We understand the requirements for all the various NAND suppliers because we work with all of them. If they choose one, two, three different NAND suppliers, we know how to work with all three of them or all two of them or all, you know. We're very good at qualifying the various NAND sources and working with them to make their product successful. We also have the highest performance and highest quality in the industry for SSD, and that's very important for these types of customers. The engagement always starts with engineering in this situation. It never starts at procurement.
The quality and the reputation Marvell walks in with is a very important asset for us in this application. Our next, I have one more testimonial here for you, and our next one will become the CEO of LITE-ON. This is Charlie Sung, and LITE-ON, if you don't know LITE-ON, they're one of the leading OEM suppliers for SSD drives, particularly in the data center, the hyperscale data center. Together, we're winning business here.
I'm Charlie Sung. I'm CEO of LITE-ON Solid State Drives Business. LITE-ON is a leading technology company, and we offer many products to our customers, like power supply, Wi-Fi module, and storage device. LITE-ON uses a unique proprietary firmware, which is configurable. This is the best and ideal choice for different customer applications. Marvell and LITE-ON have been partnering with each other for more than 15 years. We use many of Marvell's products from our solid state drive controller, and this is the best and ideal choice for different customer applications. One thing we look for in a partner is flexibility. Marvell team is always very responsive, their flexibility helps us to deliver the best product in the SSD market. Marvell has a broad portfolio of SSD controller, which enable us to address the different segmentation of the SSD market.
Today, their custom high-performance controller gives us the performance needed to win in the high-end hyperscale data center market, we are gaining the traction.
The last thing I told you in SSD, is that we're leveraging our competencies in HDD. The 20 years of experience we have in HDD, the same high quality, the same reliability that we have in HDD is going to SSD, right? Our Marvell scale, our rich IP, allows us to be more efficient also in producing SSDs. As I also mentioned before that we are doing 16 nanometers in hard disk drives. We're leveraging that same work, although we're shipping a majority of our chips today in 28 nanometers in SSD, next year we'll also be shipping customers 16-nanometer in SSD. One key point I'll make on that is the low power on an SSD drive when you put the controller in there.
These drives have a budget for how much watts they can consume. The customers tell us that Marvell's lower power solution, they put less power budget to Marvell. They can put more of the power budget to the NAND flash and run them faster. It gives them an overall performance advantage on their drives when they use the Marvell solution. Let's talk about SAM expansion. Hyperscale data centers have moved the storage from being local on the server to distributed across a data center. It allows them to be more efficient in using the memory and accessing the memory. There's limitations when you do that, right? The limitations are that you add cost because you need a front-end server to manage the storage in its new location. You have latency to the system because you have protocol translation to have to happen.
You end up essentially having memories that are slower than if they were local. It kind of defeats the purpose and the performance in order to get an efficiency of the use. NVMe over Fabric is really the future. NVMe was developed for flash, for SSD. Okay? It stands for non-volatile memory, NVM. The point here is that the NVMe will allow you to basically put your memory distributed. It looks like it's local. Okay? You have end-to-end NVMe transport of your data, lower latency. It's meant to be the same latency as being local if it was directly attached to the host by PCIe.
You basically have the best of both worlds when you use NVMe over Fabric. You get distributed data that looks local and can be accessed from multiple nodes on the network. For Marvell, this is very interesting because as the market transitions towards what we believe the NVMe over Fabric architecture, we're very well-positioned. It's not in our SAM today. We believe this will be at least a $500 million SAM. Marvell is developing now a suite of products around this architecture. Okay? The first one is we call our NVMe storage aggregator, which allows you to directly attach multiple NVMe drives to the network. It supports RAID functionality. Our NVMe storage NIC allows you to directly attach an NVMe drive right into the network.
The key point to take away from this one is that we believe this technology will be integrated into the SSD controller of the future. When that happens, Marvell is very uniquely positioned on the SSD controller to have it directly attached NVMe drives onto the network. Marvell already was first to market with NVMe interfaces for our SSD drives. We're already on multiple generations of NVMe drives. The last one is our NVMe storage processor, which really allows a seamless transition between different interfaces like SAS, SATA, NVMe. It'll enable hybrid solutions of hard disk drives and SSD drives. This actually won the Linley Award for the best embedded processor in 2016. Major achievement for Marvell. That's, by the way, on a 16-nanometer process already. That's called the ARMADA SP2. We're working with customers actively today on these products.
We're excited about the potential this market brings for Marvell. We think it's another growth opportunity for Marvell storage in the future in a data center. In summary, we are increasing our market share by growing faster than the market. We're gaining traction in the cloud. We see our mix moving that direction already, and we're seeing ramps and new design wins. We see opportunities to increase our footprint and increase our SAM in the future, which will enable new growth paths from Marvell Storage. We are positioned to grow in storage, and we're very excited about the future. Thank you.
Thanks, Dan. I'm going to talk about our networking and connectivity business, and then we're going to take a quick break before Tom and Jean come up to bring us home. My name is Chris Koopmans. I think Tom already introduced me. I've been in networking, wired and wireless, my whole career, both at startups and established companies. I think Matt also made a compelling case that our demand for bandwidth is going to continue to grow and that there's a significant amount of infrastructure required to deliver that bandwidth. Our networking connectivity products actually enable that infrastructure. Let's start with networking. In the networking industry, Marvell has a long history and is a strong player today, we actually don't think we have our fair share. I'm going to talk to you about our strategy to gain share and grow this business.
First of all, three parts of the strategy. First part is really to refocus and refresh our portfolio. I think Matt mentioned that earlier, I'm going to spend some time talking to you about what that portfolio refresh looks like and why it has been and will continue to help us grow. Second part of that is to capitalize on transitions that are happening in the marketplace. This is really required to meaningfully move the needle in this market. Finally, our entry into the data center. Very exciting space for us and we're attacking it in a meaningful way, and I'll talk about that and what it will mean for Marvell. First, a little bit of history. Matt mentioned that Marvell's networking technology goes back more than 20 years.
As he said, the very first product was actually an Ethernet PHY transceiver, which is based on the same mixed signal technology as the read channel that Dan talked about. I think even Dave Mosley mentioned it's that communication channel to turn that analog signal back into a digital signal. That's exactly what an Ethernet PHY does on a copper or optical cable. That's the same basis for our networking technology and business that was the foundation for our storage technology and business. We then, through a couple of acquisitions, got into the higher layer packet processing and software in the early 2000s. We were an early adopter of Arm and released our first Arm-based SoC in 2005 and have integrated Arm cores across our portfolio since then.
Together, this formed the foundation of a very strong Ethernet platform. We rode that platform through a significant period of growth to the point that by 2010, we were a very strong player, specifically in the enterprise infrastructure market, which is the largest segment of the networking market. At about that time, Marvell as a company started to focus more on mobile. The networking business followed suit and started to focus more on carrier. Okay. While we did gain some ground in the carrier networking market, we actually lost some of our strength in the core enterprise market, and we missed some other trends, like the initial build-out of some of the hyperscale data centers. By about 2015, there was a good recognition that this strategy wasn't working. We actually brought in a new business team to run the networking business.
Actually, it was the old business team that had grown our networking business for so long. The team immediately refocused the product line back into our core area of strength of enterprise SMB and campus. You've seen the results with four straight quarters now that we've posted a double-digit growth in our networking business. I'll touch more on data center later, but you should know that my team and I are on a mission to enter this market and take a meaningful position. First, though, before we get into data center, let's talk about enterprise. It's easy to lose track of this because the data center market is growing fast and it's what everybody's talking about. Cloud is very exciting. Enterprise still represents the largest networking market by far. It's over a $2 billion opportunity for Marvell this year.
It's also where our historical strength comes into play. If you think about enterprises and companies and the office infrastructure, there are thousands, hundreds of thousands of these companies around the world. They have legacy services that have been deployed for years. They can't just wholesale upgrade and redesign their networks every day. This two decades of hardware, software, features, capabilities is not easily replicated. You don't see all kinds of new startups entering into this space. It's a very high barrier to entry. Actually, there's only two players really with a complete Ethernet platform for enterprise today, and we're one of them. We're well established in this market. We have strength. It was really the market and the segment that the team could make the most immediate impact on our business. Let's talk a little bit about what this Ethernet platform looks like.
We have three products in the portfolio, three product lines in the portfolio. At the heart is the switch. This does all the packet processing and forwarding. The Ethernet PHY, as I mentioned before, is the interface to the analog world. The embedded CPU that runs the software, okay. There are three product lines in the portfolio. Speaking of the software, it's not a one size fits all here either. The software, there are some companies that want to take a complete turnkey solution from Marvell. We call that the Marvell Total Solution. There are others that have their own network operating system that they've been investing in for years. They need us to just provide driver support. There's a whole healthy ecosystem of third-party APIs and open source. Marvell supports all that ecosystem.
It's a very complex software ecosystem and hardware ecosystem to build into the enterprise market, which explains why there's really only a couple of players in this space today. Now, let's look a little bit deeper at the enterprise. The enterprise market is not a one-speed-fits-all market. The bottom layer is actually the access layer that connects all of our end-user computing devices into the campus network. That typically today is running at 1 gigabit per second. That moves up into the aggregation and core layers, which go to 10 gigabits and beyond. Every enterprise, large or small, still has a server room or a private data center of their own as well, and those tend to run between 1 and 10 gigabits per second today.
Within the context of the enterprise or a campus or even an SMB, there are multiple layers of networking that the OEMs who sell into this space need to provide a complete portfolio if they want to participate. Therefore, the chip providers that sell into this ecosystem also need to have a complete portfolio. When I mentioned earlier that the new business team came in in 2015, refocused back to our core market, and refreshed our portfolio, this was not just one new product. This was 25 brand-new products all delivered in the last 18 months. This was a massive R&D effort by this team to really completely refresh in the latest process technology, addressing the latest trends an entire portfolio of products. Let's look at this in a little bit more detail. On the switch side, I mentioned the two decades of experience.
This is actually our 7th generation core switch pipeline that's now been built across a variety of speeds and feeds to address the full extent of the market, from the low end, 1 and 2.5 gigabits, all the way up into the 10 and 25 gigabit core and data center products. In the PHY we've got Ethernet PHY products all the way down from 1 and 2.5 gigabit all the way up to the new 25, 50, and 100 gigabit standard products. Our physical layer product portfolio also spans the entire stack. In the CPU space, we have single, dual, and quad-core embedded Arm CPUs to address the full gamut of products in that space. We're actually the first to market with the latest Arm-based Cortex-A72 cores, which is one of the reasons why our products are being selected for this infrastructure.
As we focused on the enterprise, we've targeted specifically the needs of these products, and we're gaining design wins because of it. It's important to note that I mentioned there's only a couple of players in this space. Our competition tend to focus on the high end and sell down into the lower end. Marvell has targeted optimized solutions for the most common deployment points at the access, aggregation, and core layers for our customers, and that's why we're winning traction. Now, our refreshed portfolio is helping us, but if we want to continue that trend over the long term, we really have to align ourselves to some key industry transitions. Many people ask, "That's great that you've grown for a few quarters. Can you really move the share needle in a meaningful way over the long term? Aren't these designs done?
Hasn't this stuff been out there for a long time?" The answer is that you need to have industry transitions to help you. Okay. Things need to be happening in the industry that will actually make those OEMs redesign their equipment and make those enterprises buy new equipment and deploy new equipment in their networks. There's a couple of key trends that we're going to talk about. Number 1 is the multi-gigabit enterprise campus. I mentioned earlier that your average enterprise, SMB, you name it, today is running at about a gigabit on the access layer. Gigabit networking came out in the early 2000s, and there was a massive upgrade cycle that happened to get them all on gigabit networking. There's now clear demand to move beyond a gigabit.
Just as an example, the current shipping Wi-Fi access points actually support more than a gigabit at the radio layer. If you plug that into a gigabit access link, then your wired connection becomes the bottleneck, and you don't get to take advantage of the latest wireless access points. There's clear demand in the enterprise to move beyond a gigabit. We expect that we're just beginning the next major upgrade cycle at the access layer, and the designs for that equipment are happening now. The designs that we're winning now with our refreshed portfolio is laying the groundwork for our growth going forward, and we'll talk more detail about how that's happening. The second major trend is in the vendor ecosystem. We've seen in other markets, such as the carrier market, Chinese OEMs have stepped up and taken a significant share of that market.
Until recently, that hasn't really been the case in enterprise infrastructure. We're seeing that happening now, and we'll talk about what that means for Marvell and how we're aligned to capitalize on that trend as well. First of all, talking about the enterprise campus, if they're running on a gigabit today and there's a clear demand to go beyond a gigabit, why haven't they upgraded? The answer is the cables. Today, there's only two choices, one gigabit and 10 gigabit, and 90% of the cables deployed in enterprises today, built into the walls and ceilings and cubes and infrastructure, they don't support 10 gigabits. It's limited by physics. You can't physically get that signal through that wire. The alternative, of course, is to upgrade the cables.
That's a multibillion-dollar problem, and not to mention the construction project. Enterprises just haven't done that, right? They've just stayed at gigabit. Which is why a group of industry collaborators got together and created a new standard. In September of 2016, this new standard was ratified. It's called NBASE-T. This standard allows us to go in between one and 10 gigabits and actually deliver multiple gigabits, either two and a half or five gigabits, on the existing wired infrastructure in enterprises today. Marvell was part of this alliance, helped create this standard, and we've also developed a set of optimized solutions for enterprises who will want to adopt this technology. Once again, our competition tends to take a 10 gigabit solution where they had strength and sell it down to the two and a half and five space.
Marvell's built optimized two and a half and 5 gigabit solutions, switch, PHY, and SoC. We're winning designs because of that. We expect as this next upgrade cycle to the multi-gigabit enterprise happens, Marvell expects to gain share because of that. Let's talk a little bit about the vendor ecosystem. It's in China, there's a whole host of companies, Huawei 3Com, Raisecom, Ruijie, et cetera, that are initially, of course, targeting the enterprises in China. Chinese enterprises and cloud companies tend to be on a little bit lower end of the speed spectrum. They're not pushing the speed barrier. The OEMs that serve that market actually partner deeply and move really quickly to adopt new technology and to adopt new standards. All of this plays to Marvell's strengths.
As we've just refreshed our portfolio optimized for this space, we see those OEMs adopting our technology, and it's been a significant driver over the last four quarters. I mentioned that we've been growing. This has been a significant driver of that growth. Today, you see that these Chinese OEMs pretty much own and dominate the China enterprise market. Today, they have more than 50% share of the Chinese enterprise market. This is a story that goes beyond China. We've seen this happen before. Actually, if you go outside of North America today, 28% of the enterprise infrastructure is owned by Chinese OEMs. It's true that in North America, they don't have a significant share, but this is really not a China market story. These Chinese OEMs are gaining share for the whole world. This is a rest of world story.
You should expect that as these Chinese vendors continue to gain share in the world of enterprise networking, Marvell will gain share with them because of the designs that we have won and are winning now. In fact, let's hear from one of them now. For those of you who don't know H3C, it stands for Huawei 3Com. Initially started as a joint venture between Huawei and 3Com. Today, it's HP's division in China. They sell and market equipment in the China market. They also build equipment for the worldwide market that's sold through HP.
[Non-English content]
[Non-English content]
You heard that he said that our portfolio Well, you read anyways. That our portfolio is aligned to where H3C is going, which is important because H3C is not just a vendor to the Chinese enterprise, but actually a major vendor to the Chinese cloud companies, Alibaba, Tencent, Baidu. He also mentioned that they have adopted our campus and data center products, and those will be ramping during 2017. Let's talk about data center. I've talked about how we've refreshed our portfolio in our historic area of strength in the enterprise market, and we're going to grow share. Let's talk about the data center market. The data center market is an exciting space because it's growing fast. It's also a very diverse market, okay? It's not just one kind of data center.
Actually, even out in 2020, estimates are that the majority of ports deployed in data center switches are going to be at 10 gigabits and below. Okay, this tends to be the private enterprise data center, okay? It also tends to be the lower end of the cloud data centers and the Web 2.0 types of data centers. There's a high-end data center, which is expected to run at 25 gigabits per second and above. As Marvell looked to enter this data center market, the first place that we looked was the enterprise data center. This is the same OEM ecosystem that's buying our campus products, okay? It's the same customer base who we enjoy a relationship with, and they enjoy our technology. It's important to understand that the technology in the data center market is not so different, okay?
It's like I said, it's the same switch pipeline that we're able to adapt to multiple speeds and feeds to meet that. When we first look to enter into this market, we really want to do it in a targeted way, where we can win and make an impact in the immediate term. Okay? The first place is that 10 gigabit data center. Okay? We just this past fall announced our first terabit switch, which is designed to support 48 10 gigabit server links. This is specifically targeting that yellow half, which is our mainstream data center products. We've got design wins today that'll be ramping by the end of this year. The second place, let's look at the high-end data center. How are we going to really attack that? We know that this is a place where there's a very strong incumbent player.
We also know that there's lots of startups. Startups enter this space because it's newer, you can do it with only one or two chips, and you can get in and build some specific relationships there. This is a more vibrant ecosystem. We need to be really targeted at how we enter into the high-end data center market as well. When we looked at this, our initial approach is to find the gaps. Okay? Where in the competitive landscape today is the gap? Everybody's chasing that ultra-high-end three, six, 12 terabit switch and pushing the technology envelope, and then sell that down to the rest of the data center market. That has actually left a gap for what we think it will be a significant portion of the data centers, and that's the 25 gigabit data center.
This is where we expect a significant number of high-end data centers to be, including those Chinese cloud partners. Again, just in the last few months, we've announced our first 1.8 terabit switch, which is designed to support 48 25 gigabit server connections. This was a gap in the market. Our competitor takes their 3.2 terabit switch and sells it down to 1.8, which means we have half the power and half the size of their solution, and we're winning designs because of it, okay? This is what I mean by targeting gaps and making an immediate impact. Now, we're not ignoring the ultra-high-end either. When we first looked at the 40 gig and above part of the data center market, we looked for gaps in the solutions there that we could immediately address. For that, we focused initially on the physical layer. Okay?
All the technology that gets deployed in the data center, servers, NICs, switches, leaf, spine, top of rack, there's all kinds of physical layer products that are needed in that market as well. As an example, if you want to build a high-end 100 gigabit switch, these comes in various shapes and sizes, and quite often, you'll end up needing a PHY retimer to give the signal a boost just before it goes out the edge. Our 25, 50, and 100 gigabit retimer products are being designed into the leading high-end switches and switch OEMs today. In fact, we've started production ramps already with these products. The other thing that's happening is a move from 10 and 40 gigabit to the new 25, 50 and 100 gigabit standards.
Marvell's developed and delivered a set of gearbox solutions that'll help customers who have the products based on the old generation, to be able to very quickly adapt to the new generation. For example, today, you're going to see our solutions shipping on the world's leading server and NIC vendors with our gearbox solutions. Our physical layer products are allowing us to immediately address gaps in the high-end 40 gig and above data center market, where we can make an immediate impact. As I said, when we looked at the data center market, my team and I are on a mission to get in and gain a meaningful share of this high-growth, exciting space. We're doing it in a targeted way.
Rather than going right at the same place that everybody else is, we've initially chosen to focus on the gaps that we can address with our technology. Don't think we're stopping there. Very soon, you'll hear from us about new and innovative solutions that we think will be more disruptive to the industry, okay, which will allow us to come in and shake that market up and take a more meaningful share. Stay tuned. We have more to come. Now, let's hear from Palo Alto Networks, who's actually designed our terabit switch at the heart of their next generation security appliances.
I'm Varun Badhwar. I'm Vice President of Engineering at Palo Alto Networks. We are a next-gen security company. Our unique platform combines networking and endpoint security, along with our threat intelligence cloud, to provide protection from cyber breaches. 85% of Fortune 100 and half of Global 2000, they use our products to maintain trust in the digital world. Palo Alto Networks have been working with Marvell since 2008, almost since the beginning. We have been using Marvell switches and PHY technologies in majority of our firewall appliances. Marvell's networking solution, which integrates hardware and software, are a key front-end component of Palo Alto Networks firewall. Industry where technology is changing constantly. Marvell bring a high degree of technical expertise. Their switching product is very highly scalable and which helps us in our firewall.
As well as in every interaction with us, as well as delivering in time, Marvell has been an excellent supplier. As we have seen in the news, cyberattacks are becoming more and more frequent, and it is becoming more and more costly. Marvell, with their unique, scalable switching solutions, are helping us maintain security and trust in the digital world.
You heard him say that we have a very unique product, okay. That's reflective of where we focused on the gaps, and even though we designed it for a top of rack switch, it's finding all kinds of other homes and products because of our ability to address a gap in the marketplace. Let's summarize our networking business. Okay. We have a strong technology, rich history, and we're one of only two players with a complete platform in the enterprise space, specifically. We've refocused and refreshed our product portfolio with 25 brand-new products, and we're winning designs because of that. We are entering the data center in a meaningful way, but in a very careful and targeted way where we can make an immediate impact. Together, we expect this to drive growth.
You're seeing four straight quarters of double-digit growth for Marvell, and we expect, because of our alignment to the transitions in the enterprise market and our entry into the data center market, to be able to continue that and gain share over the long term. Let's switch gears and start to talk about wireless connectivity. This is another space that we have a very rich history. Matt mentioned we have 15+ years of organic development in wireless technology. This is a business that we're a little bit earlier in our strategy pivot. Until recently, Marvell had a very broad portfolio of wireless technologies, including mobile, IoT, module, cellular, a lot of wireless technologies. What we found is that a lot of those spaces were spaces that it was very difficult to differentiate technically.
That meant that it was difficult to get the margins that we want out of those businesses. Marvell's pivoted strategically and started to focus on the markets that we think do value differentiation, and specifically the high-performance markets that value that differentiation. What you've seen is over the last couple of years, as those lower value businesses has declined, it's left us with a much higher margin, higher fidelity, higher performance wireless business, where we now believe last quarter we've reached the bottom and we are poised for growth. Let's talk about our portfolio. Today we've got a complete portfolio of Wi-Fi and Wi-Fi Bluetooth combo solutions, starting at the low end with 802.11n client solutions. We have a whole portfolio of 802.11ac wave 2 products that are enabling the next generation of high-performance access today.
Although others are putting out press releases about technology multiple quarters before they have silicon, Marvell is very focused on the silicon development, we're very advanced in our development of our 802.11ax portfolio, we're working very closely with our customers on that as well. You should expect to hear more from us on our 11ax portfolio soon. Our customers are a testament to the fidelity of our products. Some of the top name brands that really require the high performance and reliability and quality that we're targeting. When we talk about high performance, which market cares more about performance than enterprise? Enterprise wireless is the access layer in enterprise for end user computing. It's how we all connect to the network. It's how you all are connected to the network right now.
In fact, this many people in a room, the wireless access better work. This is one of our strongest areas today. We're the leading provider of wireless access platforms for enterprises. We have the complete platform, not just the Wi-Fi radio technology, but this is that same Arm-based processor that we sell into network infrastructure, that's sold into wireless infrastructure, and it's the same Ethernet PHY that connects that to the outside world. In fact, for the wireless access points, they often are providing the Power over Ethernet as well. Our products are powering the leading enterprise access points in the world today, we have a complete platform of configurable software and hardware. How did we get there? Why are we selected as the leading vendors in this space? The reason is our performance.
We partnered early with them. We have the leading radio performance, highest concurrent user performance. We've got spectrum intelligence and beamforming solutions that allow the most efficient and highest bandwidth connections. Talking about beamforming for a second. Beamforming is an AC standard. If you have an 802.11ac wave 2 client and an 802.11ac wave 2 access point, it will allow the access point to form the beam to deliver a more efficient stream to that user. Where we are unique is we have implicit beamforming technology that is widely deployed today that allows us to, even older clients that are not based on 802.11 wave 2, we can identify them, locate them, and form the beam for them as well using our implicit beamforming technology. Which is unique, and it is important in an enterprise where you have a very diverse set of devices in the enterprise.
They are not all on the latest generation wireless, even though you have upgraded your access points. The strength of this platform continues to drive the leading edge of performance and keep us in the lead when it comes to the enterprise access point space. Tomorrow's home will look more like today's enterprise. What we are seeing is a proliferation of devices in the home. As we connect our refrigerators, and garage doors, and thermostats, and everything else, the number of users explodes. All of those are not particularly caring about the performance, meaning the bandwidth, but everything else is, as wireless becomes the distribution layer for video within the home. We have all this streaming video, tablets, and streaming video devices in the home that command ultimate performance, even in the face of growing multiple concurrent users.
Our same technology that made us a leader there is helping us to expand our business in these new high-end connected home access points and the streaming devices that provide the performance that are needed for the video streaming services. What we are seeing generally is a move where we used to have Wi-Fi as almost a given, came with your broadband connection, to now a whole new set of high-end home access points that consumers are going out and paying real money for. Rather than talk more about it, I should let one of our customers tell you more about how they are seeing that play out.
My name is Yoo- Chong Song. I am Vice President of Engineering at D-Link and Linksys. Linksys is the leading provider of wireless routers and networking equipment for homes and businesses. Today, more than 70% of the homes have five or more networking devices connected to their Wi-Fi. This is going to grow as more smart devices come online. Speed matters. Our WRT routers are blazing fast. You can basically have multiple 4K videos, play online games, and have multiple smart devices connected to the Wi-Fi without compromising performance or network connectivity. One key to delivering this consistency is our partnership with Marvell. We have been working with Marvell for more than 10 years. We use their Wi-Fi solutions as well as SoCs, PHYs, and switches on our high-performing routers.
As well as on the SMB side, we use the switches and the PHYs for the networking solutions. Marvell also works with us continuously to reflect our needs into their future roadmap that will actually help us differentiate ourselves in the market as well. That's what makes Marvell a great partner for business. Wi-Fi has become the center of our lives, we believe that we are providing the best solution for Wi-Fi to our customers. We found through our surveys that Wi-Fi has become the second most important thing in people's lives after food. Our motto is Wi-Fi for life, and people expect the connectivity to be as reliable as running water or electricity. In our industry, Linksys is a performance leader, with partnership with Marvell, we'll be able to provide great Wi-Fi solutions to homes and businesses around the world.
At least we can agree that it's important in the home and that performance will be a leader in this space, and Marvell's helping them do that. Another market that we're excited about that cares about differentiated technology is the automotive space. Until recently, wireless access has been more of a luxury in the automobile. But with some of the things that are happening now with Apple CarPlay and Android Auto and Baidu Car Life, we expect to see vehicles to adopt wireless access as a key part of the technology throughout the product line. This is also a market that values differentiation.
If you think about the relative value of Wi-Fi to a Barbie house or something like that versus a car that's going to be built over a period of a decade and supported over a period of another decade, the level of quality of technology in the latter is far more important than in the former. This is a space that we've targeted early. We've also found that it's a space that allows us to differentiate. If you think for a moment about what happens inside a car that needs to support screen sharing for Apple CarPlay and a Wi-Fi hotspot for the screens in the back seat and streaming Bluetooth audio or hands-free calling all at the same time, all from within one chip, it's a complex solution.
Marvell invested early in this space and has a unique multi-antenna architecture that creates enough separation between the antennas that we're able to deliver very low interference and high performance solution that can do all of this at the same time. This is why we're getting adopted in cars today. Although today what's actually shipping is a small number of vehicles, we're actually number one in the automotive Wi-Fi space. We have all the customers to show that, they're using our technology, we're currently designing into the next generation of vehicles. We also see 802.11p as an adjacency in this space. We're sampling products in that today. What this will allow you to do is to have the vehicles communicate with other things, vehicles to vehicles, and generally V2X communication.
This is a new standard that's being developed, which we think, if widely adopted, will really drive wireless solutions throughout the connected car of tomorrow and throughout the entire product line of these automotive companies. As you start to see that proliferate, you should expect to see Marvell grow with that. We covered a lot of ground. We talked about our networking portfolio, where we have a lot of strength, and we've refreshed and refocused our portfolio to grow in enterprise by taking share. How we're entering the data center market in a meaningful way, but in a very careful and targeted way. How we've repositioned our wireless products to a market of high performance and high reliability where we can get the value for the differentiated technology that we build.
Finally, we're expanding into new markets that we think will demand these types of high-quality solutions from us going forward. Together, this is helping Marvell, and this is what we think is going to let us grow going forward. Now, let's take a quick
I'll get it.
Okay.
Yeah.
Hopefully, we still get a break.
We're going to take a 10-minute break. We're running a little bit behind, so be back in here by 11:35 A.M. latest. Jean and I will bring you home. Coffee and restrooms back through the hallway back there. Thank you. Okay, start heading to your seats. I want to get you guys out into the snow as quickly as I can. I am. I'm sort of goosing them back slowly. You guys want to start working your way back to your seats? We'll get started again. I'm herding them back. We're getting them back. I want to get you guys out into the snow as quickly as I can. Let's get started, guys. Let's get this thing going so we can get it done.
As many of you know, it's a bit unusual to have sales and marketing presented at one of these investor conferences. Normally, you want to hear from Jean and Matt, Dan and Chris. You want to understand the product strategy and direction. You normally hear from them. As a matter of fact, not only do I not normally present at these sort of conferences, I don't normally attend them. I'm usually out meeting with the people who are going to drive our revenue, drive our profits. As far as market cap is concerned, that's Jean's department, and I'm not normally here. These are not normal times, okay? Last year, this company went through a fairly dramatic series of events that left a lot of our investors and our stakeholders wondering, did we still have the customer lock?
Were we still engaged with the customer, or did we create a lot of irreparable damage that we couldn't recover from? That's why we thought, okay, for this conference, it's probably legitimate for sales and marketing to give you an idea about our relationships with our customers and what we're doing in the sales side to actually make sure that the strategy you heard put forth this morning is successful in the customer community. That's why I'm presenting today. First of all, you heard from a few of our customers this morning. As a matter of fact, you'll hear from seven of them today, and you heard it in English as a first language, English as a second language, you even heard it in Mandarin. It sort of speaks to the global footprint that we've actually created at Marvell.
We cast a fairly broad net when we started the idea of customer testimonials. Seven of the eight people we talked to responded positively. The eighth one wanted to actually do a testimonial with us, but logistically, we couldn't make it work. Seven of the eight actually responded positively. These are some of the things that you actually heard this morning. Marvell has great technology. Marvell goes above and beyond. Marvell is a trusted partner. Marvell stands the test of time. That should certainly speak to any concern that may be out there about whether we maintained our customer lock, and whether we have it going forward. Okay? I know a lot of you out there, I know a lot of you are thinking, "Why did you join Marvell?
You had to be fairly comfortable coming out of your last gig. You're getting kind of old for this worldwide sales stuff and another global gig. Aren't you getting a little too old, fat, and ugly to do this kind of stuff?" I figured in order to answer that question, I'm going to tell you a little story, okay? First of all, as Matt said, he and I have known each other for a long time, and when he was named to the CEO of Marvell, I sent him a text message, and that resulted in a series of texts and conversations that culminated with Matt saying, "Why don't you consider joining us and helping us get this thing back on the right track?" I said, "All right, let me take a look. Let me look under the hood." This is what I found.
First of all, when I looked into the technology toy box, it was chock-full. World-class technology and technologists inside this company. When the team looked last year at how are we going to reduce our focus down to the markets that are going to drive growth and return the greatest amount of value for our shareholders, there was plenty to choose from, okay? That certainly was a very pleasant surprise. The second thing I saw was as we focused our resources down onto our chosen markets, we de-emphasized R&D investment and watched our expenses in other markets and improved our operational efficiencies, our revenue started generating profit. We started generating cash from operations. What that actually gave us the ability to add cash to an already healthy cash balance of $1.7 billion and no debt to service.
I said, "Oh, geez, profitable, strong balance sheet." I started looking at the customers and the customer relationships. Now you heard from six so far, you'll hear from seven today. When I went to CES this year with these guys, we had some 200 meetings in four days, and I've also met with a lot of our largest customers, and what I'm hearing consistently from them is two things. Thing number one is as the semiconductor industry consolidates down, they are paring their supplier base. They're reducing their supplier base, and they're focusing on creating deeper, stronger relationships with fewer suppliers. The second thing they're saying is, "Marvell, you are a vital part of our supply base going forward." This is a pretty strong statement. Our customers have a vested interest in our success.
When I looked at this and said, "Geez, world-class engineering, great technology, profitable revenue, and a strong balance sheet, and customers with a vested interest in our success," the question should have been, "Why wouldn't I join Marvell?" This is an exciting time for the company. We're going to have a lot of fun getting this thing back on track and back on a growth trajectory. Now, there's an old saying out there that says, "If you want to catch big fish, you got to fish where the big fish are." The first order of business as we get there is we have to align this sales organization to catch the big fish. Now, the funny story about this slide was our agency when they first said, "Oh, you can't show this. This is sport fishing. It's no longer politically correct. You're going to offend your audience.
The picketers up at Trump Tower are going to move down to the St. Regis. They're going to protest your presentation." I got to thinking about it and I said, "Well, this is Wall Street. These are investors. These are meat eaters. They're not going to get offended by this slide." We left the tuna slide in just for you guys. You're going to need a bigger boat. Yeah, going to need a bigger boat. The first order of business when I got into this thing was I got to align the sales organization for maximum success in our target markets with our target customers. What does that mean? First thing you got to do is you got to look at the coverage map. Where are the big fish? We got to make sure that we're aligned with them.
You've seen there's been a series of reductions at Marvell. The last one, I actually reduced the sales organization. I had to because we had people in areas of the world where there were no longer target customers. I'm replacing them with people in the areas of the world where our desired and target customers sit. With the sales strategy, we aligned our most valuable resources in those areas where the target customers were. We use a hybrid sales approach. I have sales reps, and we're actually restructuring the sales rep organization. We're adding some mega reps for broader coverage and greater focus.
We're adding reps that are targeted at selling networking product lines, we're augmenting that with global distribution, we're re-examining the entire global distribution network, and we're going to be reducing our distribution network, getting it more targeted, more focused on our target markets. We're training our sales organization on how to manage large accounts, large account management techniques, strategic selling techniques, giving them all the tools, the collateral, and things necessary to increase their close rates at the customers. The last thing we have to do is we have to improve the awareness for Marvell. We have to improve the brand and brand equity.
You're going to see a lot of activity in marketing around these chosen markets in the coming year, outbound marketing campaigns and events like this, conferences, press releases, even a social media campaign to actually improve the awareness of Marvell in our target markets. Finally, we're going to be spending a lot of time throughout the customer chain, from our customer to our customer's customer, all the way down to the end user, making sure that that entire value chain is aware of and has a preference for Marvell Technology so that when our customers go to sell, their customers are saying, "What about this Marvell stuff?" Creating a greater awareness of our technology. The first order of business, get this sales organization aligned. That's what we're doing. We're aligning them on these markets.
I already told you, we're putting our highest value resources near these customers in these markets. We're embarking on an effort to upgrade the field technical resources near these customers, bringing the product knowledge and education closer to the end customer so that we can close in a more efficient manner. We're moving to single opportunity system across the company so everybody knows where the large opportunities are, who's involved, and what their actions are. We're doing an extensive outbound marketing campaign. The first thing that I did this year when I looked at where we were spending money and on what conferences, I whacked about five of them because these were conferences in markets that just don't service our strategic direction going forward. Most recent was Mobile World Congress in Barcelona.
By not going to that conference, I was able to save the company about $150,000 and redirect a chunk of it into the Open Compute Platform forum that went on in Santa Clara this week, where it was chock full of customers from these industries. We were meeting with the likes of Facebook and Cisco and Dell and HP and Arista. We had a keynote address at that conference, another speaking engagement. We were actually able to focus down with that gold sponsorship on a bunch of very valuable meetings that were targeted on our customers that we want to do business with in the data center and in the cloud. That's what we're doing as we go forward, to be a lot more focused and targeted on our messaging and on our strategy.
Now, I'm a sales guy, so now that I'm done with that message, I have to talk out of the other side of my mouth, and tell you about adjacencies and optionality in our markets. We spent a lot of time talking about the center of that picture today and how we need to widen the pipes because there's so much information being created around the periphery, around the network edge that's coming into that network. We've talked about the consumer IoT. There's a lot of activity over there from virtual reality gaming and augmented reality and PCs and connecting your thermostats and printers and toasters to the internet. That's all well and good, but there's a whole segment over here on the right that I like to call commercial IoT.
These are the markets where massive amounts of data are being created on the edge of the network, processed, and sent back to some aggregation point, and there's a whole phenomenon in there that we like to call edge computing. We want to talk a little bit about that because all of the products that are being created to make those pipes faster and perform higher are also able to service these markets. You think about a smart city, for example. You've got a Wi-Fi mesh of interconnected cameras tracking the movement of people and vehicles through the city. High-end gateways bringing massive amounts of disparate data in, ripping it apart, and routing it to the right place. Smart cars. The amount of data created by an autonomous driving car is phenomenal. It would just stagger the imagination. Oil and gas exploration.
Anything from seismic analysis to drilling to maintaining the rig creates massive amounts of data in an area of the world more likely where they have a very low bandwidth connection back to the cloud, so it has to be managed locally. All the way down to automated factories. These factories that have machines that do automated processing and measure the milk fat in cheese or the roundness of a ball bearing. They get this information back, they take an action on this information, and then they send it back to an aggregation point. Every one of these things is doing edge computing and sending data back to that cloud. Let's talk about one of those. We talked a lot about automotive today. Car industry is going through rapid transformation. There are three disruptive trends: connectivity, autonomous driving, and new energy.
These trends are driving an explosion in sensors and data collection. The cars that are out there today, the ones that are being driven around today, they have CAN buses. These are 1990s technology buses. They're all separate. They're all disparate. They take care of whatever part of the car they need to take care of. The problem with the CAN bus was it was a pseudo-standard, few suppliers, it had a limited lifetime. The car company said, "We want a single standards-based, scalable data backbone that could scale with their needs." That was the birth of automotive Ethernet. That happened a few years back, and there's more and more and more traction around taking Ethernet as the single backbone to the car and replacing all of these disparate buses.
We are already engaged with and have wins with many car companies in 100 megabit and gigabit, buys and switches today. If you think about it, you add this move to automotive Ethernet to our established automotive Wi-Fi business, and Chris just talked at the very end about our position in the automotive Wi-Fi. We're well-positioned for data movement, and if you're going to be moving all this data around the car, it's got to go somewhere. There's a movement to bring our flash control into the automobiles. If you look at all the products that were created for that cloud, Wi-Fi, switch, processor, and flash control, they're all finding their way into the car, leveraging that R&D investment into an adjacent market. Now because the car is changing so greatly, the automobile electronics are going through a sea change of transition.
New technologies like self-driving cars, over-the-air updates, subscription-based applications, really does turn this vehicle into a server on wheels, as Dan had said earlier. That continues to drive the proliferation of sensors and the data that needs to be stored. The interesting thing is, I saw the other day a commercial for Toyota Corolla. The Toyota Corolla is now adding collision avoidance as a new feature for next year's car. The Toyota Corolla. This is the bottom line of their platforms. If you look at it, we're certainly not all that surprised by the proliferation of intelligence and autonomy on the high-end platforms. What's surprising to us is how rapidly this technology is driving to the base platforms in these vehicles. This is, in fact, creating an even larger market for us at a more rapid place than we had even imagined.
This is one of the market adjacencies. If you think about the edge compute system, every one of these things, and I don't care whether it's Wi-Fi connected cameras or automated machines or smart cities, self-driving car, they all do the same thing. They sense data, they put that data in memory, they process and filter that data, and they send it somewhere else. Every one of our products that's being developed for that core market, that little product in the middle, can be sold into this market. Our multi-core CPUs, our flash controller, and any one of these devices to take the data off of the controller, Ethernet, NVMe, Wi-Fi.
As a matter of fact, I've got one of our distributors in Asia designing this board because they have a whole plethora of suppliers who put these products into automated factories and write software for it. I said, "Well, if they're going to write software for automated factories, they may as well do it on a development kit that we create." We're actually creating almost an exact version of that product to actually send into the automated factory market. I told you what my first order of business was, get the sales force lined up to be successful in our target market. Okay? Second order of business is I'm going to take all of that intellectual property that's being created for a target market, and we're going to drive it, storage, networking, connectivity, into those adjacent markets and increase our SAM by billions of dollars.
That's our second order of business as the sales organization. If I can get you to take anything away from my talk today, I would say, first, we're actively and aggressively restructuring this sales organization. I hope to be done with most of the moves by the middle of the year, and then we just basically continue to increase and drive the efficiencies and competencies in this organization. We want to put our most talented, our most valuable resources out there so we can actually go after the big fish. Second, more fundamentally, we have a customer base with a vested interest in our success. They're working with us. They're actually collaborating with us on roadmaps. They're making sure we're doing products that they can buy going into the future. They want us to be successful. Third, we're going to get our message out.
I already told you, we're going to be targeted in a very aggressive manner this year around the data center, around the cloud, and getting the message out through a lot of outbound marketing campaigns and through a lot of business development work with the customer supply chain all the way to the end. Fourth, we have an amazing set of technology. We're going to leverage that technology into that expanded SAM to drive additional growth. As I had said earlier, the fact that we have world-class technology and technologists, profitable revenue and a healthy balance sheet, and customers that want us to be successful, has created a phenomenal platform for us to launch into future growth. That's why I actually came to this company, and this is going to be a lot of fun driving forward. Don't listen to me.
I'm going to let our last customer today, one of our largest customers, tell you about what we're doing. We'll let you see from Rubik at Western Digital.
I'm Rubik Hovsepyan, Chief Procurement Officer of Western Digital. Western Digital is a leading provider of storage technologies and solutions. We provide hard disk and flash disk-based technologies for consumer, enterprise, and cloud-based environment worldwide. We have been working with Marvell over 20 years. We have been using Marvell storage controllers and read channels in over 2 billion of Western Digital products. Recently, we are using Marvell preamps in our latest two and a half inch client solutions. Marvell defines what it means to be a true partner. They work with us on roadmap, design, and qualification. Marvell does what it takes to make Western Digital successful. Marvell has been a pioneer in re-channel. As such, they continue to invest and develop technologies that help with time to market and power efficiency and overall system performance.
Marvell's taking innovative approaches in making investment in new technologies such as convergence of storage and networking. As such, through the innovation and investment technology, we believe Marvell's committed and a good partner to be in this business with us. We are in an era where data is creating a world that's more predictive, more productive, and more personal. It enables creativity and discoveries and enables deeper connectivity. Western Digital is enabling the world with new ways of utilizing and leveraging data. We have been the largest Marvell customer in several years in a row. With what's in the pipeline, we anticipate to remain a top customer for years to come. Together with Marvell, we will continue to power the technology and create solutions that will deliver the possibilities of data.
Okay. My name is Jean Hu. I'm CFO of Marvell. I joined Marvell last August, it has been a very exciting and fun journey. I'd like to take time right now to thank everyone for spending time with us today and your interest in Marvell. This morning, what you have heard from our team is Matt talk about our focus and strategy in storage, networking, and connectivity, and the tremendous opportunity ahead of Marvell. You also have heard from Dan and Chris to talk about why Marvell is uniquely positioned to create value to drive top line revenue growth. Tom just talk about the building go-to-market capabilities and expand and drive our revenue growth opportunity. You also have seen our new leadership team.
The longest tenure actually is just nine months, but we have achieved a lot and did a lot together as a team. As a team, when we think about creating shareholder value, the way we think about it is we want to build a strong business to generate top-line revenue growth, and also generate a consistent and sustainable long-term cash flow. As a company, we are uniquely positioned. We have three drivers that we can leverage to create value for shareholders. The first is revenue growth, the second is margin expansion. The third is we actually have a very strong balance sheet and a very strong financial flexibility. That's another lever we can delever to create value for shareholders. I'll spend the next few minutes to cover three topics.
First, I'll do a quick recap of our restructuring plan, and also talk about fiscal 2018, the current fiscal year. Second, I'll tie together what you heard this morning from our team about the top-line revenue growth opportunities and what it means to our long-term top-line revenue profile. I'll also talk about the margin expansion opportunities Marvell has and what it means for our long-term financial model. I'm sure that you're all waiting for that. The last is the capital allocation, which we have a very strong financial flexibility. To deploy capital to create value for shareholders, that's another important lever. Okay. First, let me do a quick recap of our restructuring plan. When Matt joined last July, we were just finishing the first half of our fiscal 2017. At that time, we reported a gross margin for the first half to be about 53.6%.
The operating expense run rate, annualized run rate, was $1.1 billion. Matt mentioned earlier, the team got together, we went through a rigorous strategy and portfolio review. Three months later, last November, we announced our restructuring plan. Here's what we said last November. We said we would expect our gross margin to get to 58%-60% in the second half of fiscal 2018. We also said we're going to cut our operating expense by $240 million-$260 million to get to an annualized run rate of operating expense of $220 million-$240 million exit Q4 of fiscal 2018. I'm very pleased to update you. Last week, when we reported our Q4 fiscal 2017 earnings result, we guided the Q1 fiscal 2018 gross margin to be approximately 59%.
We now see we'll get to 59%-60% gross margin in the first half of fiscal 2018. On the operating expense side, last week we said we are one quarter ahead of our restructuring plan. Now we actually see we'll narrow our operating expense run rate to $820 million-$830 million, exiting Q3 fiscal 2018. As you can see, our team is executing extremely well, and we are ahead of our plan. Now I'd like to provide you some perspective on fiscal 2018. Typically, we don't talk about fiscal year comments, provide the fiscal year comments. We're right in the middle of restructuring actions. There are a few dynamics that we do want to take you through. Matt talked about it earlier, is when we cut our operating expense, it's about 25% of the company's operating expense. It's a very large cut.
It's about $250 million. When we look at the revenue associated with most of the operating expense, it's actually either consumer focused or it did not yield return on investment. We classified all those revenue under this other category. Those revenue will decline in fiscal 2018. As a result, we expect the revenue from other category to decline approximately 30% in fiscal 2018. On the other side, if you look at the area we're investing, which is storage, networking, and connectivity, all three business are going to grow in fiscal 2018 and more than offset the decline in the other category. This is the revenue picture we're thinking about in fiscal 2018. Now let's look at the profitability. As I mentioned earlier, we are ahead of plan.
When we execute on our gross margin improvement plan and operating reduction plan, you can see we're going to see very significant margin expansion in fiscal 2018. Over the year, if you look at the contribution from gross margin expansion to our bottom line and operating expense reduction to our bottom line, we believe we expect to grow our earnings per share by approximately 60% in fiscal 2018. It's very significant. As a company, when you look at fiscal 2018, it's really a year Marvell is building a strong business model. It's a year of transformation and building a foundation, and from this foundation, we are positioned very well to grow Marvell. Now let me shift gear to the long-term view of the company. First, on the revenue growth.
You have heard Matt talking about fundamental demand drivers to drive the market we participated to grow much faster than overall semiconductor market. You also have heard Dan and Chris talk about our strategy to grow the business much faster than the market we participate. I want to talk about storage first. I think we all agree storage market overall is growing because the demand for storage capacity is growing much faster than the supply we can deliver. You heard Dan talk about it, we are really agnostic to where the storage market goes, because we're one of the very few companies who has the depth and breadth of technology, IP portfolio, and the engineering expertise to address every storage market, no matter where it goes. That is actually why we are the market leader in HDD market, and also the market leader in SSD market.
Dan also talked about it, we're the only company who can provide all the solutions in different segment of SSD market, which as you know, it's very dynamic market and there are a lot of innovation going on. Let's talk about the planning assumption. When we think about the financial model, I look at a spreadsheet just like you guys every day. On the HDD side seriously, our planning assumption is really we hold a very cautious view about HDD market. We expect HDD market actually to decline low to middle single digit in next several years. Accordingly, we actually think SSD market is going to grow around 15% CAGR in next several years. In HDD market, you heard Dan talk about we're going to expand our market position in nearline and cloud data center to gain incremental revenue.
Secondly, we're investing in the preamp opportunity, even though in fiscal 2018, we don't expect meaningful revenue from this opportunity. In the longer term, we do strongly believe we'll gain a fair share of market because of the technology leadership and the market positioning in HDD market. When we look at both the incremental revenue from preamp and also the opportunity in the cloud and the nearline, we can offset the overall HDD market decline to keep our HDD business largely flat. On the SSD side, it's really exciting for Marvell. Dan talked about it, we already got the revenue to be more than 20% of storage revenue in Q4 fiscal 2017. When you look at the opportunity, we are participating literally every segment of SSD market. For us, we're very confident we can grow our SSD business much faster than the overall SSD market.
I'm pretty sure, you ask me all the time is, okay, how about HDD market decline much faster than the planning assumption we're using. I would just say, if that's the case, it has to be SSD market is growing much faster than 15% we're planning, or there is some new technology in the market that fill the gap, because we all violently agree the storage capacity is increasing much faster than we can deliver. From that angle, when you think about Marvell, really, we are agnostic to where storage goes. We have the solution and the technology leadership to address the innovation, the dynamics of the market to be the leader and to gain share in this market. We're very confident we can grow our storage business faster than the storage market.
We can model, I model all the way from between the mix of HDD and SSD. The important takeaway is we're agnostic and we can address all the opportunity. Now look at the networking. Chris talked about our refreshed portfolio of 25 products in SoC switch and the transceiver side. It's very important because for us, this whole portfolio literally is driving a unique product cycle for Marvell. We're literally at the beginning of this product cycle. Fiscal 2017, our networking revenue grew 13% year-over-year. When you go through the product cycle with us, not only the design wins we already have won is going to ramp into revenue.
I don't know if you guys noticed, because I'm Chinese, the H3C guy actually said, "Oh, the new product will ramp in 2017." At the same time, it's also true is we continue to get strong tractions in networking market and winning designs with all the customers, Chris talked about. When you look at the networking market, it's growing 6%-7% and our opportunity to grow much faster than market is absolutely very, very near and dear to us, and we think we can deliver that. On the wireless connectivity, this is the business, as Chris mentioned, we're literally just going through the transition and just completed the transition to get out the low margin Wi-Fi business. In Q4 fiscal 2017, we said the wireless connectivity business has bottomed.
Going forward, really, when you look at the segment we're focusing on, either it's enterprise access or automotive market. We are both number one in those two markets. The newly connected home market, we are getting strong attraction because of the performance of our wireless portfolio and the technology. I really like what they said, is Wi-Fi is second to food. I cannot agree more with that. For me, Wi-Fi actually is number one, because without wireless access, the food, I don't care that much. I think it's very important, is Wi-Fi market is growing very significantly. I think Marvell will be there, to address this market growth opportunity. When you put it all together, our core business will grow faster than the market that we participate. By the way, I did not include what Tom talked.
I insist I don't include anything Tom talked about it, the expansion opportunities. For instance, Ethernet automotive. We actually have very strong design wins and strong tractions with all the major customers. If those design wins become revenue in fiscal 2020, that will be upside to our model here. We're not including the current view, how we want to help you to think about Marvell financial model. For modeling purpose, we're modeling other product lines to decline 10% or less. For some of you may remember, Marvell actually has one of the nicer portfolios of printer business and the customer ASIC business. Those businesses are not growing, but they're actually not declining either. Marvell probably is the only printer, silicon ASIC supplier left, but that market is going to last very long time.
This other product line is actually going to decline, getting to be a very smaller pie, as Matt mentioned earlier. It's probably one point to offset the overall company's growth rate. That's on the revenue side. Let me talk about margin expansion next. On gross margin side, this is really a focus of the company from Matt, the whole team, and our head of operation. If you look at what we have done is, really there are two drivers on gross margin. One is the product mix, the other is the cost reduction. On the product mix side, you know we're getting out all the low-margin consumer business. That will improve our product mix. If we look at our fiscal 2018 model, we believe that will contribute approximately 40% of gross margin improvement. On the cost reduction side, our team has done a fantastic job.
The head of operation and the whole operation team really did great on supply chain management across all sides of the supply chain. We actually are expecting close to more 60% of gross margin improvement coming from the supply chain management side. It's an effort for last nine months, but the full benefit of the cost reduction started to factor into the overall Marvell model. After we get to this 59%-60%, we certainly have an objective to continue to drive the gross margin higher. Really, there are two things we need to do. One is the product mix. What we believe is that when we launch new product, when we ramp up the new product into the product line, because we have Tom as our head of sales, I think the gross margin will be better for all the new product we launch.
That will help us to continue to improve gross margin. On the supply chain side, we're going to continue to improve our supply chain management and efficiency. That will be more comparable to all the other semiconductor company are doing. You can tell, don't measure it by the way, you can tell the green box going forward on supply chain management side will not be as significant as we're doing currently. Overall, we are going to improve our gross margin to be greater than 60%. Some of you probably are going to ask is, what do you mean greater than 60%? Is it 61% or 62% or 63%? I think just hold your thought. I want to talk about in the whole context of our long-term financial model. Next, on the operational excellence side, operating expense management.
Matt and the team is really building a culture inside the company to really build a machine, execution machine to focus on execution and the discipline. We want to be a very disciplined operator of our business. What we're doing is, we actually are implementing a very rigorous R&D review process to track, measure, and also manage our R&D investment so we can improve productivity significantly. Also, we're centralizing the procurement functions. As Matt mentioned earlier, this company actually did not have a centralized procurement function. When you think about how much money we spend as a company, this is unbelievable. Again, it's about tracking, measuring, and also managing the overall spending of the company. The other thing we're doing is that we're simplifying organization layer and the structure. Those take time, we'll get there. It's all about the efficiency of the company.
In the longer term, we are building infrastructure and the processes so we can really grow the business and support the scale going forward. We're automating the process too, try to establish a shared service center to further reduce cost to improve efficiency. Of course, the other thing is about the site consolidation. During this current restructuring process, we have consolidated our sites significantly, there's always more to do because Marvell, as some of you know, we actually own a lot of buildings we occupy in different countries. Our view is when we consolidate the site, some of the foreign site, if we don't need the office building we own, we'll monetize them. Of course, our largest site is our U.S. campus. We also own the whole real estate there.
Our view is we always want to figure out a way to see financially what makes the most sense for shareholders and for the company. Over time, that's a longer term target, we're going to continue to optimize our site strategy, how we can increase the density of a site operating more efficiently as a company. Overall, we're targeting our R&D as a percentage of revenue to be 22%-24% in the longer term. We are very determined to target our SG&A percentage of sales to be around 8%. Let me get all together, to take you through how we think about our long-term financial model. From fiscal 2018 to fiscal 2020, we expect our top line revenue, the core business, is to grow faster than the market that we participate.
On the gross margin, we expect to be greater than 60% during this period of time. Our R&D, as a percentage of revenue, we are targeting 22%-24%, and the SG&A, as a percentage of revenue, we're targeting 8%. Of course, our objective is to get to approximately 30% operating margin. Marvell actually has a very efficient asset-light business model. Our capital spending is about 1%-2% of revenue. The beauty of this model is our free cash flow as a percentage of revenue, actually, is very close to our operating margin. This is a very powerful and compelling financial model. It's about, in the end, all of you guys know, it's about a sustainable, consistent free cash flow. That's the value of the business. Let me shift to our capital allocation. We had about $1.7 billion cash in end of the fiscal 2017.
With the business model, especially the free cash flow percentage as the revenue is approximately 30%, our business is going to generate very significant cash flow going forward. That's a tremendous financial flexibility. The way our team think about capital allocation is really a balanced and a disciplined approach. A balanced approach means we're going to focus on investment as our number 1 priority, organically or through acquisition. At the same time, we're very committed to return cash to shareholders. We're going to keep our dividend payment and use buyback to return cash to shareholders. As you know, our board of directors authorized a $1 billion buyback plan last November, and since then, we have been actively buying back our shares. In the longer term, we are implementing a policy to return at least 50% of free cash flow to shareholders.
That's the way we think about in the longer term. We do want to make that kind of a commitment as Marvell's policy going forward. Let me summarize. As a team, we're really executing well and try to get the restructuring and complete that on time. We are building a very compelling long-term financial model. The last, we have a strong financial flexibility, and we're going to utilize it to deploy capital, either create more value or return those cash to shareholders. With that, I'll turn to Matt for summary.
All right. Thank you, Jean. I thought you did a fantastic job. Everybody get excited by all that? Everybody ready for lunch soon? Okay, great. Again, Jean, great job, and also the rest of my team today, and all the Marvell team members that actually worked really hard to pull this day together, I want to say thanks. As you can see, I'll just end with some quick closing notes, and then we'll do a little Q&A, and then I'll break for lunch where you can have some more informal discussions with myself and the team. Basically, look, we're very excited, as you can probably tell from all the presentations and our body language and our emotions With the future prospects of this company. We've got great technology. We've got great customer partnerships. I think you heard it directly from the customers themselves.
Overall, I want to reiterate my belief, which is that Marvell can grow organically and participate in the growth rates of the markets that we're servicing. It's an exciting time to be involved with Marvell. It's great to have you all here. With that, I'm going to have my team join me on stage, and we'll do some Q&A. Then we'll go grab some lunch. Okay, come on up.
Are we on? Okay. We're going to try and do about 10 minutes of Q&A. There'll be microphones in the audience. You guys know the rules, who you are, where you're from, who the question is directed at. If you don't do that, we reserve the right to not answer your question. How's that for a legal disclaimer, Mitch? I see Mitch has approved of that. Oh yeah. Right there.
Thanks for taking the question, and thanks for all the information today. Craig Ellis at B. Riley, appreciate it. The question is regarding the 6% growth rate. Is the intent to be growing at that rate in fiscal 2019 Matt, or is that after a period of investment in some of the products and SAM expansion initiatives that were outlined by some of the presenters? Thank you.
Sure. Yeah. Happy to answer that question. I think the way we think about it is, again, tried today to paint a picture of the markets that we're in and what growth rates those are at. As Jean alluded to, we've got this period of time we're going through restructuring, where we've got this other product roll-off core growing actually quite nicely. Overall, we think of that as sort of a post-restructuring growth rate that we aspire to attain. Again, if you heard today with the product portfolio and the lineup we have, that's a growth rate that we'd like to see ourselves in, again, post restructuring for 2019 and after.
Yeah. Right here.
Still up here. Let's get a mic up here closer, please.
Right here.
Yeah. Okay.
Yeah, go ahead. There's one back there.
Yeah, it's John Pitzer with Credit Suisse. Thanks again for all the information. I guess my question, two quick ones for Dan on the storage front. One, when you think about hard drives kind of being flat, to what extent does that include the preamp? Who's supplying the preamp today? Who do you expect to take market share from? Just a broader question on storage. A lot of people have made the point that you're kind of agnostic to the medium. I want to challenge that a little bit. If you think about your sort of a $ per exabyte or gigabyte of storage, HDD versus SSD, it seems like actually the $ content on SSDs per volume of storage is significantly higher for you.
Why wouldn't sort of the HDD to SSD transition that a lot of people are worried about actually play as a strong growth driver once you get through kind of that initial transition of HDD being the vast majority of the market today?
Answer the first part, and I'll take the second.
Yeah. Repeat the first part one more time for me, please.
The HDD guidance for flat-
The preamp. Okay
to what extent, how much is that as TAM versus market share growth?
Yeah. We weren't really including the preamp in what we showed today, right? We showed the kind of maintaining is, I guess that's the point, right, Jean? Without the preamp in there.
Yeah. I think if you look at the HDD market, the overall market, our planning assumption is to decline about 4%.
About 4%, right.
When you include preamp, the preamp market actually is growing slightly. It's about 3%.
It's about 3% up, right.
When we look at our incremental revenue opportunity in preamp, beyond the fiscal 2018, it's a long cycle product. Also the market share gain in the cloud and the near line, when you add those two together, that's how we look at the HDD market, and now can offset that 4% or 5% decline of overall market. It's between both, right? The share gain in near line and the price cloud, also the preamp incremental opportunity. I think to just answer your preamp question is, this is the market. There are only two players, primarily two players in the market. One of the incumbent, they don't have the HDD business anymore. For us, we have HDD business. It's a natural leverage, like Dan talked about it. Yeah.
Yeah. That's good.
The second part I'll take just because I think you pinged me on this, I've actually had a chance to think about it where he hasn't. I'd make one statement before I answer that, though, is just as a tone issue here, I think as you've all seen today, hopefully, you picked up on something, which is we took a very modest, conservative, thoughtful view of how we presented our business today. The preamp was one example. We actually consciously decided, let's not put that in there as the SAM to make it look better. We just said, "Hey, that's upside if we get it, so let's take it out." Same thing actually on automotive, right. We know for sure we've got a very strong automotive design win pipeline. I've been personally engaged with those customers. We feel good about that business.
It's not a near-term growth driver, and so we wanted the investment community today to really get a good, clear picture of the Marvell today and have a mind to not oversell in the future, although we're very excited about it. Just as a tone issue, some of your questions may be around that. On your question specifically, which was really, hey, if you sell an HDD controller into a terabyte drive, let's say, and then you sell an SSD controller into a flash drive that has a much lower memory capacity and it's the same ASP, aren't you getting more ASP per byte? You're absolutely right. That's absolutely a fact, and it's true. We intentionally, to my point, did not model that and sort of try to even make that claim at this juncture. That's a true statement.
You could actually argue as SSD grows and we're agnostic and we're with the right customers, there's actually an opportunity there. It's not modeled today for the reasons that I mentioned.
Richard?
Timothy Arcuri, Cowen. Thank you. I guess my question is on M&A and sort of how you balance your comments about capital return. You're going to generate $750 million a year roughly per the model, of which you're going to return half, so you'll have a couple hundred million per year. You have some real estate, you have some cash, you don't have any debt. The focus is really in networking, but the assets that you could buy in networking are pretty big. I guess I'm wondering how you think about balancing taking on debt onto the balance sheet, levering up to do something more sizable in networking. Thanks.
Sure. Yeah. I'll answer the M&A question. I'll have Jean add to it. It's a common one, right? It's really how does that play into our capital return thoughts and policy, and I think as I stated when we even announced the billion-dollar buyback, which was, hey, we want to try to have a balance of returning cash to shareholders, but maintaining flexibility and some firepower if we want to go off and pursue some M&A, especially on things that are close to what we do. We're still in that mode of maintaining flexibility.
We're not mature yet enough in our journey to kind of be, and I think it was you or somebody else that framed it as, "Hey, are you a TI?" Which was return it all, or are you a Broadcom, which is aggressively deploy it towards M&A, we're still trying to strike that balance. We're also trying to strike that balance in the context of a restructuring plan that we bit off that was quite substantial and that actually we're driving ahead of schedule. To get all that executed is keeping us busy. Those are the dynamics that you should think about. The last thing I'd add is that as we've gotten more settled into the company and I think you've now gotten a preview more out of the black box of what the assets of the company look like.
You can imagine that Marvell is really a platform technology company with a lot of scale, and our ability to actually capture synergies if we were to go off and pursue M&A, actually, I think is meaningful because of the current size and portfolio of assets that we have in the company. Can't answer it exactly head-on, but I'm trying to give you a framework of how we think about it today being seven to eight months into our journey here at Marvell.
Right here. He's been trying for a long time.
Let's just keep a mic closer. Sarah, why don't you come up here?
Okay.
Thanks. I take the question. Christopher Hemmelgarn for Barclays. In terms of who answered that, I guess I'll narrow it down. It's probably not for Chris. You laid out a really interesting case for why you can grow. When I look at the hard disk business, specifically, share gains were a big component of that staying flat. Just realistically, this is a market that's been pretty well established. You guys used to talk about share gains back and forth all the time. It never really moved all that much. Could you just talk about what gives you the confidence that this time's different?
I'll give a quick one. I'll let you guys if you want to add to it. I'd characterize that to be very clear that we think of this as very modest share gains, okay? Again, this is not a management team that we don't come out. We have not beat our chest today. I think we've been very thoughtful about our comments and how we think about things. You're right, that's a market that has two established very good players. The proof points that we would point to were even in our most recent quarter in our guide, where we actually bucked seasonality because we saw some nice growth on new platforms and new programs at customers and those types of drives.
Again, we're not out here to say these are going to be massive shifts. When you think about a market that's declining a little bit each year, if you can move the needle up a little bit each year, for us, that segment of the market commands better gross margins. It's a higher premium, tougher segment. It plays to our strengths. You get the preamp thing going. That's how we get to our model. We recognize it's two very good players. Customers want choice. We're being thoughtful about it. Did you guys want to add anything?
No, that's.
I think that's perfect. Okay.
Please.
Hi. Christopher Rolland, Susquehanna. Congrats on the Analyst Day. I'm still trying to figure out what company this Analyst Day is representing. It says Marvell behind you guys, so I guess I'm going to go with that. Congrats on Marvell 2.0 so far. I wanted to talk about SSD controllers. I remember back in the day when it was kind of a two-horse race. It was you guys and SandForce. SandForce was sold all around and eventually went to Seagate, where they kind of de-emphasized that market, and maybe some new players stepped into that void. Marvell didn't really capitalize on that opportunity. Maybe you could talk about what happened there, why that didn't happen, and why it's kind of reinvigorated for you guys now and those opportunities there.
Yeah, go for it.
Okay. Well, none of us up here were actually here during the time of the SandForce Marvell battle, but I do recall it, and I'm aware of it from our internal reviews. I think Marvell has had its own SSD challenges in the past. I think some of you recall that, especially exposure to one large PC OEM a few years ago, where that business was really concentrated. I think actually, quietly behind the scenes over the last couple of years, especially, Marvell actually has been taking advantage of this opportunity. If you see the momentum in our business, and a lot of you have been able to.
able to kind of back into what our growth rates look like based on the commentary we've given. This business has grown incredibly fast, and as Dan pointed out, it's not a one-trick pony anymore, right? It's really a business that's highly diversified across the whole range from retail all the way up to the highest of high-end and hyperscale data centers. We actually think we've benefited from it. There's probably some other companies that benefited as well from that particular dynamic you mentioned about SandForce. We like our prospects and how we're positioned, and we've got great leading indicators in each of that segment of SSD now.
Harsh? We'll let you bring us home, and then we'll break.
Thank you. Harsh Kumar, Stephens. Question for Matt, and maybe Jean as well. I think you guys mentioned that you started the supply chain initiative about nine months ago. It typically takes about nine months to one year to get benefits. I was surprised that you feel that you've already gotten a bunch of the benefits. How much do you feel you got out of it so far, and what's in the margin, and what do you think is left? I've got one more.
Yeah. When you think about our fiscal 2018 gross margin improvement, as I said, probably 60% of that improvement is from supply chain management. The effort has been ongoing for the last nine months. Over the fiscal 2018, you're absolutely right. It takes six to nine months for the effort to get the full benefit. The supply chain management, the margin improvement, that's why you are going to see in the second half of fiscal 2018, we're going to see the whole thing. Right now, when you look at our guidance of 59%, certainly part of that is because the supply chain management the team has done six months and nine months ago.
Okay. As a follow-up, I think your R&D today is running about 29%-30%. I think you said long-term goal is 22%-24%. What are the obvious things? Is it all leverage from revenue growth? Is that the assumption, or are there obvious things you can take out?
Yeah, I'd just say I think it's a combination of two things. One is, as you saw, we've done better than expected on OpEx through the restructuring. We've pulled it in by a quarter, we've narrowed the range, that's a positive, right? If you just think about directionally how things are going. The second is, as you've seen in our growth plans today, we believe we can grow this company modestly. With some modest revenue growth and the trajectory we're on with respect to how our OpEx is sort of rolling into our restructuring, we believe we can get there, and our internal math works. We gave a range, right? Obviously. We're still focused on making sure we have an efficient company, that we're organized properly. I mean, all those things you shouldn't assume that we've just sort of stopped, just like on supply chain.
Just because we got some benefit, we haven't stopped. This is a continuous improvement mentality in the company. Last point I'd add, just back to your first question. The one thing that surprised us but was a benefit was operations team's done a great job on the supply chain management, which does take time to take benefit. You can imagine there was a lot of low-hanging fruit given where we were at. Some of the benefit we saw up front was just literally getting in there and unpacking the cost and figuring out what we could do differently, and I think we executed pretty well on that. I think we got kind of a twofold benefit. One was just attacking it, and the second was just getting some systemic fundamental improvements to it.
Okay, we have lunch out there in the place where we had breakfast this morning. We'll all be available to meet with you guys, and we can continue the conversations. Thank you.
Yeah. I'm around as long as you guys want. Anyway, great to see all of you. Thank you.
Thank you so much.