Welcome to Broadcom Inc.'s third quarter fiscal year 2018 financial results conference call. At this time, for opening remarks and introductions, I would like to turn the call over to Tom Krause, Chief Financial Officer of Broadcom Inc. Please go ahead, sir.
Thank you, operator, and good afternoon, everyone. Joining me today is Hock Tan, President and CEO of Broadcom. Today, Hock is going to give you a detailed review on our core business and spend some time outlining the industrial logic behind our recently announced acquisition of CA. I will spend time reviewing our Q3 results and Q4 outlook, and most importantly, our financial model and capital allocation policy. Quickly on the formalities, today's call will primarily refer to non-GAAP financial results. A reconciliation to US GAAP measures is included in today's press release, which is available in the investors section of our website at broadcom.com. Information on risks that could cause actual results to differ materially from the forward-looking statements made on this call is also available in today's press release and in our recent SEC filings. This conference call is being webcast live.
A recording will be available via telephone playback for one week and archived in the investors section of our website at broadcom.com. At this time, I would like to turn the call over to Hock. Hock?
Well, thank you, Tom. The strength of our business model delivered another quarter of very sustained revenues, strong earnings, and free cash flows. Consolidated net revenue for the third quarter was $5.07 billion, 13% increase from a year ago. EPS came in at $4.98, a 21% increase from a year ago. While free cash flow at $2.13 billion is 42% of revenues. We have a lot to cover today, let's dive right into the segments. Starting with Wired. In the third quarter, Wired revenue was $2.3 billion, growing 4% year-on-year. This segment represented 45% of our total revenue. Third quarter wired results reflect strong year-on-year growth for both our networking and compute offload businesses, driven by robust demand from the cloud data center markets, as well as traditional enterprises.
Networking and compute offload represented approximately 60% of our total wired segment in the quarter and grew over 10% year-over-year in the quarter. This is off the back of growing over 15% annually in the second quarter. This part of the wired segment is doing really well. However, cyclical headwinds in certain parts of our broadband businesses have impacted year-over-year growth for the wired segment. While digital subscriber line, or DSL demand remains stable, demand for PON, fiber to the home in China, as well as video access, particularly in North America, has been soft compared to a very strong 2017. As a result, broadband was down year-over-year in the third quarter after being down in Q2 as well.
Turning to the fourth quarter fiscal 2018, we expect networking and compute offload to continue to grow double digits year-over-year as strong demand from both the cloud and traditional enterprise sustain. However, cyclical headwinds we have seen in video access, including cable and satellite, are persisting into the fourth quarter. As a result, in the fourth quarter, we expect Wired to grow only mid-single digit year-over-year. On the other hand, we are very encouraged by the prospect for fiscal 2019. We expect strong growth in our networking business to continue, driven by new product ramps of our Tomahawk 3 switch and our Jericho 2 router platforms. We also continue to see strength from our deep learning ASICs with our cloud customers. We forecast broadband video will bottom this fourth quarter as we start to enter an upcycle in 2019.
On enterprise wireless access, too, we expect to be the first to enable integrated 802.11ax chipsets during this coming year among service providers, enterprises, and homes. Let me now turn to Enterprise Storage. For the third quarter 2018, Enterprise Storage revenue was $1.25 billion, representing 25% of revenue. As we had experienced and mentioned in our wired networking businesses, robust enterprise IT spending drove over a 70% year-over-year revenue increase. This, of course, includes contribution from Brocade, but even without Brocade, storage was robust year-over-year in the third quarter. Looking at the fourth quarter, strong demand from enterprise continues to be good. We expect year-over-year storage revenue growth to accelerate. Moving on now to wireless. In the third quarter, wireless revenue was $1.3 billion, which was flat year-over-year. The wireless segment represented about 25% of our total revenue.
In aggregate, wireless revenue were in line with our expectations for the third quarter. We benefited from the initial seasonal ramp at our North American OEM customer, which was partially offset by anticipated decline in our other large wireless customer. We expect this ramp at our North American OEM customer to drive wireless revenue to be over 25% sequentially, even as it may be down single digit year-over-year. Let me take some time to put this in perspective. Like all our franchises, our RF frontend business, which makes up roughly half of our wireless segment, competes, and competes very well based on its technology, leadership, and its ability to deliver differentiated high performance products generation after every generation. To generate the high returns we expect on our substantial R&D and manufacturing investments, we focus on delivering the best FBAR technology in every new generation of smartphones.
Nonetheless, from time to time, not that often, but time to time, the same technology platform used by our customer may extend beyond one generation. When this happens, it does create an opportunity for our customer to temporarily use lower performance alternatives in selected SKUs. With the benefit of hindsight, this may be precisely what happened with this 2018 generation. Every indication we have is that the cadence of annual platform upgrades will resume in the upcoming 2019 smartphone generation, and we believe we are very well positioned to win back the platform. With 5G on the horizon, we expect this cadence of annual upgrades to sustain. As a result, we are maintaining our high level of investment as the market transitions to 5G. Meanwhile, in Wi-Fi, Bluetooth, the transition to 802.11ax continues to keep us in the lead.
We believe we are very well positioned to sustain this particular franchise over the next several years. Accordingly, we expect to see our wireless revenue returning to double-digit growth in fiscal 2020, following a temporary dip in fiscal 2019. Finally, our last segment, industrial. In the third quarter, the industrial segment represented 5% of total revenues. Excluding IP licensing, the industrial business was up over 10% year-on-year. Distribution resales continued to be strong with double-digit Q3 year-on-year growth. We expect the demand environment for industrial to remain strong and industrial resales to maintain double-digit year-on-year growth during the fourth quarter. In summary, we continue to execute well on our business model. More than half our consolidated revenue, you may note, is benefiting from strong cloud and enterprise data center spending.
This, coupled with a seasonal uptick in wireless, will drive our forecast revenue in the fourth quarter to be $5.4 billion, an increase of 11% from a year ago. In the meantime, our margins continue to expand due to our focus on technology leadership and high performance products. This is all driving exceptional cash flows, which provides us great flexibility in our capital allocation model of returning cash to shareholders through dividends and share repurchases, while enabling us to pursue strategic acquisitions to expand our earnings capacity going forward. Speaking of acquisitions, before I turn this call back to Tom to talk about the financials in greater detail, let me perhaps take a few more minutes and talk about CA Technologies. The number one question we get with CA is why did we choose to buy it? Cut to the chase.
We're buying CA because of their customers and their importance to these customers. CA sells mission-critical software to virtually all of the world's largest enterprises. These are global leaders in key verticals, including financial services, telecoms, insurance, healthcare, and retail. CA does it at a scale fairly unique to the infrastructure software space. This can only come from long-standing relationships with these customers that span several decades. In other words, these guys are deeply embedded. Broadcom does a lot of business with the cloud companies building the digital economy. The leaders, Google, Amazon, Microsoft, are all large customers for us. They are growing rapidly, and we are, as you notice, growing with them. They use our leading-edge silicon solutions to develop their next-generation data centers to enable many businesses worldwide.
On the other hand, when you look at the largest enterprises, which comprise CA key customers, these guys really have limited direct access to our mission-critical technology. In that lies what we think is a new and huge opportunity. Just as we have done with hyper-cloud players, we believe we can bring our compute offload solutions, our Tomahawk switches, Jericho routers, fiber optics, and our server storage connectivity portfolio directly to these same large enterprises that are buying CA software. These large end users invest tens of billions of dollars on IT infrastructure every year. Through CA, we believe we have a big doorway to engage strategically with these customers and provide them direct access at very compelling economics to the same leading-edge networking, storage, and compute technology that are used to enable the cloud service providers today.
Beyond this industrial logic, I might note CA by itself is a great franchise. Mainframes remain the backbone of the enterprise computing environment and are relied on to run mission-critical applications. Mainframes process approximately 30 billion transactions per day and $7 trillion of credit card payments annually. Contrary to popular belief, over the last 10 years, mainframe workloads have actually increased 3.5 times, driven largely by increasing amounts of data generated with every single transaction. Given mainframes power the most important parts of larger enterprises, we believe this will remain a strong and stable market opportunity for us for a long time. CA is a leader in delivering a suite of mainframe solutions across application development and ITOM tools. Bottom line, we actually see this opportunity, a great opportunity, I might say, to double down for future growth.
With that, let me turn the call over to Tom at this time.
Thank you, Hock, and good afternoon, everyone. My comments today will focus primarily on our non-GAAP results from continuing operations, unless otherwise specifically noted. Let me walk through our results for the third quarter of fiscal 2018. Third quarter net revenue was $5.07 billion, just ahead of the midpoint of our guidance. Our gross margin from continuing operations was at the high end of our guidance at 67.3% as we benefited from a more favorable product mix in the quarter. Operating expenses were slightly lower than we expected at $874 million, driven by lower SG&A. As a result, operating income from continuing operations for the quarter was $2.54 billion and represented 50.1% of net revenue. Adjusted EBITDA for the quarter was $2.71 billion and represented 53.4% of net revenue.
For housekeeping purposes, Q3 depreciation was $129 million in the quarter. Below the line, net interest expense was slightly better than guidance due to higher interest income from our cash deposits. The tax provision was in line at 7% of operating income from continuing operations, so $170 million. Excuse me. The diluted share count was 453 million shares and includes the weighted average impact of the stock repurchases completed in the quarter. As a result, the company delivered $4.98 of EPS in the quarter. This represents 21% year-on-year growth, including the impact of share repurchases. Working capital, excluding cash and cash equivalents, increased approximately $209 million compared to the prior quarter, due primarily to an increase in receivables.
This increase was driven by seasonally higher shipments in the last month of the quarter, as well as the effect of a distributor consolidation program for the Brocade business, where we are providing a temporary extension of payment terms to facilitate the consolidation. In addition, cash restructuring expenses were $18 million, as we are now at the tail end of the Brocade integration. Finally, we spent $120 million on capital expenditures, which was slightly below expectations. As a result, free cash flow from operations was $2.13 billion, or 42% of revenue. This represents 52% year-over-year growth in free cash flow from operations. In the quarter, we returned $754 million in the form of cash dividends and spent $5.38 billion repurchasing 24 million AVGO shares. We did not pay down any debt in the quarter.
We ended the quarter with $4.14 billion of cash, $17.6 billion of total debt, and 438 million fully diluted shares outstanding. Now let me turn to our non-GAAP guidance for the fourth quarter of fiscal year 2018. This guidance reflects our current assessment of business conditions, and we do not intend to update this guidance. This guidance is for results from continuing operations only. Net revenue is expected to be $5.4 billion, ±$75 million. Gross margin is expected to be 67%, plus or minus one percentage point. Operating expenses are estimated to be approximately [$874 million]. Tax provision is forecasted to be approximately 7%. Net interest expense and other is expected to be approximately $125 million. The diluted share count forecast is for 436 million shares. CapEx will be approximately $110 million.
Before we open the call for questions, I want to update you on our financial model and capital allocation policy. On the financial model, there have been a lot of questions regarding our long-term growth and concerns about the growth rate of our core semiconductor business. The intention with the CA announcement has not been to signal a change in the growth rate of our core business. On the contrary, we believe our long-term growth rate for the semiconductor segment will remain mid-single digits, driven by end market growth and content increases from new product introductions. As we acquire businesses outside of semiconductors, including Brocade and more recently, CA, we are taking a conservative approach relative to our internal expectations on revenue growth. The returns we model do not require growth to hit our targets, but make no mistake, we do expect to grow these businesses.
The important message is that we do not see any fundamental changes in our long-term growth rate. Now, on to capital allocation. Here at Broadcom, we have a set of complementary, highly profitable technology franchises that require limited capital expenditures and that sit on top of an efficient corporate platform. This, in turn, spits out a substantial and sustainable base of cash flows that we expect will grow over time. This expected cash flow generation provides us with a lot of flexibility on how we allocate capital to create value for you, the shareholders. We are committed to our policy of distributing 50% of our prior fiscal year free cash flow to shareholders in the form of cash dividends. Given our fourth quarter outlook and expected full fiscal year 2018 results, we anticipate another substantial increase in the quarterly dividend for calendar 2019.
With the remaining free cash flow, we see the opportunity to do a couple of things. One, we plan to continue to buy back shares. We currently have $6.3 billion left on our $12 billion stock repurchase authorization that extends through the end of FY 2019. Two, with a focus on maintaining our investment-grade credit rating, we believe we also have the cash flow and the borrowing capacity to continue to expand our earnings base through strategic and accretive acquisitions. Finally, as previously announced, we have cleared HSR with respect to the CA transaction in July. The transaction is still subject to CA shareholder approval and antitrust approvals in the EU and Japan. We expect to close in the fourth calendar quarter of 2018. That concludes my prepared remarks. During the Q&A portion of today's call, we request that you limit yourselves to one question each.
With that, operator, could you please open up the call for questions?
Certainly. Ladies and gentlemen, if you have a question at this time, please press the star, then the number one key on your telephone keypad. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. Again, in the interest of time, we do ask that you please limit yourself to one question before rejoining the queue for any follow-up questions. To prevent any background noise, we also ask that you please place your line on mute once your question has been stated. Thank you. Our first question will come from the line of Timothy Arcuri with UBS. Your line is open.
Hi, this is Pradeep Ramani on behalf of Timothy Arcuri. I had a question more on the longer-term view of CA and the Wireless Solutions Group and how you view
The Wireless Solutions Group, from a strategic standpoint, given that the CA acquisition is kind of focusing you guys towards more of an infrastructure company.
That's a very interesting question, it affords me the opportunity to clarify how we look at our, as Tom calls it, set of businesses. Our business model is very much focused on putting together a portfolio of what we consider technology product franchises. That's not necessarily limited to IT infrastructure or networking or data centers in any particular specifics. As you notice, we have a range of products that sells into multiple end markets, which ranges from wired, and even in wired, we make a distinction, as I said, of networking, data centers, as well as more service providers spending as it relates to carrier access and video delivery. That's two sets of end markets by itself. We have enterprise storage, which is very data center centric, I mean.
You're right, we have wireless, which as we define it, is very focused on mobility on smartphones, where we put out the best, latest technology. Finally, we have industrial, where there are set of products that goes to various industrial products, not necessarily connected to data centers. They are very disparate, they are very diverse, and therein, in our view, lies our strength. It's a very set of diverse product franchises, and that's the key to operative word, franchises. Each of them has a set of common characteristics. One is they operate in niche markets, typically. Those are the niche markets that become mass markets because mass markets have moved over to these niche markets. Part two is we have the technology, modern technology.
We are the leader in technology in each of these niche markets, we tend to have the highest market share too, in each of these end markets. The common thing we do is they all sit on a platform, but each of those niches keep investing, and you have seen the level of dollars we invest. We are investing over $3 billion, $3.2 billion a year, just on R&D and product development as we move to each product generation and evolves the technology for use of end customer. We make sure we lead in each of those. We still believe, to answer your question specifically on wireless, we believe our position in those wireless products and markets, that we are very much in the lead technology. We are lead in market share in the niches we are in.
It is satisfying those considerations of franchises in those specific markets as you would apply to switching and routing in data centers where we are very well represented, too. The benefit of all these particular franchises is they all are enormously profitable, and they all continue to grow.
Thank you. Our next question will come from the line of Pierre Ferragu with New Street Research. Your line is open.
Hey, Hock and Tom. Thanks a lot for taking my question. Let us hear about Computer Associates as well. You have demonstrated in the past a rather unmatched ability to create value from your acquisitions, and this is something you have mostly done in the semiconductor industry, and that is what you got a lot of investors and analysts used to. Today, it really feels like we need to better understand the Broadcom model, how Broadcom can create so much value from acquisition and how it can apply to CA. In that spirit, my question to the two of you, Hock and Tom, would be, can you tell us what you guys do like nobody else?
What makes you unique at integrating a business you acquire and create value from these businesses like nobody else, and in particular, like a private equity fund, for instance, would not be able to do? Then, of course, put that in the context of CA Technologies. How are you going to apply these unique capabilities to CA Technologies?
Thank you. Interesting question. Let me try to address that if I could. To start off with, we are not private equity. By no means are we. Why? We understand the businesses that operates in Broadcom very well, and we operate those businesses. We are not financial investors. The financial performance, the capital allocation that comes out, the exceptional cash flow we generate out those, making those businesses very successful happens to be just the end product. We run those businesses, and we run them as a group. That's the biggest difference between us and private equity. Very much so.
Where we see some differentiating traits in how we identify and acquire those businesses and then integrate them into a whole as part of Broadcom, it's simply this: I think we're very aware of our ecosystem, what product lines, what markets are very sustainable, very good as potential profit and growth opportunities. We're very focused, and that's the key thing. We are very focused on determining what businesses make sense to invest in and what businesses we do not or should not invest in because it won't generate a return. Which is why, to expand on my reply to the earlier question, today Broadcom comprises 19 separate product divisions, each of them leader in their own right in each of those niche markets they are in. By being extremely focused on continuing to be the technology and market leader, which basically means delivering generation after generation.
Because in fact, one advantage in technology is you keep having to evolve with better and better products that your customers can use and ask for. As you do that, you actually create more and more value to your customers. The extension of that is shown by the fact that if you look at our financials over the last several years, we expand our product margin, gross margin, as a collective whole 100 basis points approximately every year. This is the same product. This is not about adding new acquisitions which accrete the gross margin. We're talking about if you look organically at the same products that we have three, four, five years ago, you will see that margins expand as a whole. The reason it can expand is because you're delivering more value to your customers.
The real basic thing is be very focused, stick to what you do very well and focus on where you are very successful and keep doing the same thing. What we do when we look at acquisitions, very simply, is we look at companies where the opposite is actually sometimes happening, where the core business of the company or some of these business companies are not so focused on or neglected in many ways. Instead, the bright, shiny objects get focused on where the strength of that particular company may not be so good. We basically pull them back to their roots and put them into as part of the overall Broadcom portfolio. That's really what we have.
Thank you. Our next question comes from the line of William Stein with SunTrust. Your line is open.
Great. Thanks for taking my question. Hock, one of the biggest questions that we've gotten, especially more recently, is on the semi cycle. Now, I understand you have 19 or so franchises that you could argue are more specialized, but I think you're certainly exposed to the broader trends in the industry. There's an expectation that we're seeing a slowdown, in particular in China, especially potentially related to tariffs. I'm wondering if you can offer a comment as to where you think we are in that cycle, what you see going on in that regard. Thank you.
That's a very good question and very timely question. I'm not trying to look out far, nor trying to basically postulate a vision here, but what we're seeing now and what we've seen recently, and looking what we're seeing now, is that the dynamics of the semiconductor space is constantly changing. I know that's an obvious answer, but what I mean is by different end markets. We, in some ways, are fortunate in selling to four or five end markets, very different end markets. I can tell you over the last two years, the behavior of all those five end markets have been very different. It's hard for me to say how's the whole semiconductor industry, because it does cover into a lot of spaces.
Two, like one or two years ago, I did say that in 2016, 2017, even 2017 and 2016, broadband was very strong. What part of the, I guess, level of service provider spending worldwide, but also leading to business that kind of cyclical. It's a business on my end that's relatively flat, but sustainable and cyclical. Today, as I say, broadband, as I mentioned, is not so strong anymore. Now, last year, two years ago, data center spending was still okay, but this year, 2018, is extremely strong and continues to look good. We see different parts of the cycle, just like even wireless. Wireless two years ago, three years ago, was great. Content was growing. What we've seen over the past year is smartphones, literally, not just handsets worldwide, but smartphones just kind of flatten out, totally flatten out.
Where cost becomes a concern, but more than that, demand on innovation becomes limited, and people are now waiting for perhaps a 5G cycle before we see another update. Industrial, oh, yeah, automotive was moving away for a few years, drives industrial. Continues to do so, as we see, though we start to see definitely some slowdown from where we are, both in automotive and industrial. You're seeing ups and downs across different segments, different end markets that some of which uses semiconductors. I guess our best saving grace here is because we are fairly diversified, we kind of keep ourselves stable and secure on a total basis, as opposed to riding any particular end market upwards or downwards.
Thank you. Our next question comes from the line of John Pitzer with Credit Suisse. Your line is open.
Yeah, good afternoon, guys. Congratulations on the strong results. Hock, maybe the short way to ask my question is, does this operating margin expansion story have a ceiling at some point? I guess the longer term or longer way to ask the question is, I wonder if you could just talk a little bit about how you think about R&D. I think oftentimes investors get fixated on R&D as a % of revenue and forget that at your scale, your absolute dollar spend is just enormous. Help me understand, is there something about your IP portfolio that gives it more leverage than a typical digital or SoC company or why are you able to drive so much more leverage out of your R&D line than many of your large peers?
As you answer that question, maybe you can talk about, is there a ceiling to this op margin?
That's a great question, John, I'll try to address it. You're right, it starts with IP. We sell intellectual property, except we productize it. A lot of our business are semiconductors, and we sell IP embedded in silicon. It's perhaps a simple best way to describe. That's what we do. We don't try to license this out. We make it into products that addresses what our end users, end customers need to use or further make it into part of a systems. It's that intellectual property that drives the technology evolution because we keep feeding that machine. We keep enhancing, innovating on those technologies in any particular markets we're in. As I mentioned, we have 19 of those markets. In each of them, we behave very commonly.
We have a team of people, and in many cases, we probably in all, most cases, I would add, we have the best engineers in the world, architects and engineers in the world, in each area, in the space they are in. We are among the best. These guys is the other side to IP, takes the IP they have developed over the years and innovate to the next better thing. We keep doing that. Customers love to have this product because it makes them successful, makes them more productive, it makes them do things that otherwise they can't do. When you do that with each evolving technology, each evolving generation of technology, you basically get a higher value added to your product. Always do because you give your customer more value, you get something more for it. I give you an example.
The well-known Tomahawk 3 switch is a 12.8 terabit per second throughput switch. The previous generation, Tomahawk 2, was only 6.4, half that throughput. You are able to put into a data center and on top of a rack of servers twice the throughput, twice the capacity, twice the bandwidth. Am I able to charge 2x the price? Of course, no. That's not the way technology works. We are able to obviously improve again, our value simply by the fact that even as dollars, our price per terabit drops on a total 2x terabit, the value of the product goes up for us, for the customer. If the demand, the usage, consumption increases to use up all 12.8 terabit, basically, the data center scale gets to scale out tremendously at a very cost-effective set of economics.
That's an example that applies across all our 19 product lines. To boil down to at the bottom, at the end of it all, because of that, our ability to do that, to offer better products or more value to our customers, our product margin goes up, and it goes up faster than the amount of R&D we pour in, and we pour in quite a bit, to sustain that level of improvement. That leads to an improved operating and expanding operating margin. That's what we've seen. That's what has happened.
Thank you. Our next question comes from the line of Ross Seymore with Deutsche Bank. Your line is open.
Hi, thanks for letting me ask a question. Hock, I wanted to focus on your wired business, both in the near term and the long term. In the near term, you did a great job of explaining some of the puts and takes between that 60/40 split of the fast-growing and slower-growing businesses. To the extent it's 45% of your business today, and we look forward longer term, how should we think about the growth rates of that 60/40? What does it mean to the profitability of the company, either on the gross or operating margin line, as the growth rates seem to be so different between those two sub-segments?
Well, Ross, thanks for asking that too. What you say is true right now. In 2016, I loved broadband. It was on an upcycle. You recall there was this Summer Olympics floating around. Everybody was signing up for cable, digital cable access. We were booming. That time, hey, that thing outperforms data centers. Networking, that is. Today, the cycle turn around. We look at broadband and say, "Man, this thing is dragging me." It's not. When the upcycle happens, as you fully expect within the next 12 months or so, it'll be great again. That's one of the interesting thing about it in looking at, say, even wired, even wireless. A lot of these markets has their ups and downs, especially if you look at it quarterly, much less annually. Quarterly, even worse.
What we have to keep realizing is these are all technology-driven applications, end market-driven, as better and better, more innovative technology comes in. It keeps expanding, some at a slower pace than others. What it does is it adopts new technology, which allows us to keep adding more value as we progress through it, even though it goes through its ups and downs. The key thing to all this is sustainability. These are all very sustainable end markets. The product we see today may be much better than the product in this end market we saw five years ago, and the product five years from now will be much better than product we see today. Believe me, the end use continues.
Thank you. Our next question comes from the line of Craig Hettenbach with Morgan Stanley. Your line is open.
Yes, thank you. Hock, I have a question for wireless RF. You play mostly at the high end of the market. Interesting developments in China when I think of some of the local brands there, some impressive specs coming out and as they start to ramp to volume. Just want to know if that potentially could change your opportunity set in RF in the China market over time.
A very interesting question, it is, again, part of the whole franchise model. That, yes, because again, in the case of wireless, we have very unique technology. We have very differentiated technology that allows us to produce very high-performance products in those smartphones. So far, it's been the super high-end smartphones that tends to use our products. I could see a situation where, especially with 5G coming into the mix, with all these difficult bands showing up, where 5G, you start running higher than three gigahertz spectral bandwidth, you start to need better and better RF components, especially in filtering, very much so in filtering. We could see that being required across the board in many next phase or next generation 5G phones.
We can see that happening, in which case then it starts to expand beyond just high-end phones, and you are exactly right in that regard. It has happened before, a few years ago, when there was certain bands that were so critical, it can only be done using FBAR. Very difficult FBAR technology. For one, that was pretty cool.
Thank you. Our next question comes from the line of Blayne Curtis with Barclays. Your line is open.
Hey, thanks for taking my question. I just want to follow up on that wireless. You were talking about cycles and some years better than others. When you look at wireless, cellular, and then Wi-Fi as two components, just kind of curious, as you look out, you mentioned 5G, what's the content story between now and a couple of years from now when 5G will be the majority? Then on the Wi-Fi side, you just talked about timing of a
Okay. It's actually two aspects you're asking here, let me try to address them separately. Wi-Fi, in many ways, is a more stable, predictable evolution of the technology trends. Today, very much so, everybody uses the Wi-Fi standard called 802.11ac. C as in China. That's great. It's definitely an improvement from what it used to be five years or 10 years ago. We have a new technology, new protocol coming in called 802.11ax, which I commented on a couple of times in my prepared remarks. What ax does is, in a nutshell, it increases, in layman's language, the bandwidth. Huge. You can imagine easily running data stream wirelessly from your handset upwards and uplink and downlink way over one gigabit per second, even two gigabits per second, you can tear it away.
What's even more interesting is it allows for multiple users simultaneously, which is something that's always been tricky. It requires a lot of technology, hardware, and software. We are in the lead in doing it, as I indicated. We are the first out with our products. They're working, we're designing, and we start launching it with multiple partners starting October, that's next month by the way, this year. Starting with the retail routers and going on to the enterprise access points, and then operators by early next year. It's a big thing, 802.11ax. I bet you in the spring, you will find at least one big handset maker coming out big time to push 802.11ax, and our chips will be right in those flagship phones. It's more predictable.
Our technology is so strong, I have to say that we see ourselves in the roadmap of our key customers over the next two, three years. More predictable as 802.11ax starts to go to a second wave, an upgrade, and all that. On RF FBAR, it's truly not that much different, except that what's happening here more than anything else is over every several years, we go from, as you know, 2G, GSM before to 3G, and then we had 4G that's been going on now for six years. Now there is a demand, or kind of a demand I might add, for even higher speeds, higher throughputs, lower latencies, more connections. That's what 5G is all about, which leads to IoT applications and all those applications we had dreamed about previously. Those are great.
It's just that those are very, very difficult to implement, and in order to make it happen, one, in a nutshell, it will happen as it has happened in the past on 4G, is your phone already runs 3G, 4G. You now have to put in additional components, additional capability to run an additional set of spectrum that is operating in 5G. Once you start doing that, you really have issues of coexistence. You also have to reach out in 5G to frequencies that are much higher, much more difficult to produce, to put in a phone for data communications. I'm referring to frequencies that go way above 3 gigahertz now, as a first step. As we go into more and more of these 5G phones, you have more frequencies, more requirements or components in the same limited space of a phone.
You start to have to innovate on your components to be able to fit them in a phone, a limited space, lower power, working very well together. That's where our capabilities, our technology in RF, especially FBAR, comes into it all. That's why we see this as a sustainable franchise, especially for someone who is able to design, have the IP to design capabilities, components that few people are able to replicate.
Thank you. Our next question comes from the line of Chris Caso with Raymond James. Your line is open.
Yes. Thank you. Just a follow-up question with regard to the wireless and some of the prepared comments that you made. You talked about expectations for return to double-digit wireless growth next year. You also talked about being in a position to win back some of the FBAR business. Can you reconcile those two comments? Is one dependent on the other for the double-digit growth you're expecting next year? What are the assumptions behind that?
Chris, it's Tom. Let me just clarify the prepared remarks. What we were referring to is double-digit growth in 2020 off the back of a dip in 2019. Just to also clarify, we do feel really good about the prospects of winning back the business that we discussed that we had lost in the current generation phone, which would impact the very back half of 2019, but really would have an influence on the 2020 growth rates we discussed.
Thank you. Our next question comes from the line of Harsh Kumar with Piper Jaffray. Your line is open.
Yeah. Hey, guys. Thank you. First of all, congratulations. Great numbers. Thank you for the clarity on CA. The big question we're getting is: What is Broadcom's expertise in running a software company? This is the first one for you guys in this area. Could you maybe talk about some of the strengths and challenges you see and maybe some of the plan around running this business? Also for the non-mainframe business, what kind of margin opportunity do you expect to see from, for example, enterprise solutions?
Well, I love this question. I'm almost tempted to tell you, well, the same reason we put together a bunch of businesses in semiconductor, what you call semiconductor solutions. On one extreme in our semiconductor solutions, I have plain, simple hardware, semiconductors, pure hardware, analog components. To the extreme, it's not even silicon in some cases. It's nanotechnology. It's indium phosphide, gallium arsenide, as in lasers. To the more well-known, well-recognized silicon SoCs, silicon on a chip that we build our routers and switches and deep learning chips on with loads of software, by the way, on this thing. Lots of software. In our networking teams, I have as many software engineers as I have hardware engineers, silicon solution engineers.
In our video delivery business, video which is basically a set-top box, the cable modem, I have more software engineers because there are a ton of different kinds of software that goes into a set-top box chip compared to hardware engineers. We understand software. CA, you're right, is all software, but set-top box is 60, if I have to put a number, over 60% software. If I have to look at a switch, I can make a switch that is simple, programmable, and it's called software-defined networks and write a lot of software stacks to program this, and that's 70% software, 30% hardware, or I could hard code chips as I do in certain other versions of my networking business which are typically lower-end switches. I would say I have 80% hardware and 20%, 30% software. It could cross a lot of spectrum.
One thing is common, it's technology. It's technology solutions you provide to your customers who cannot operate very productively, very efficiently, sometimes operate at all without that. It's technology solutions that evolve over time and your ability to keep up with customer needs over time. We're very good at managing that. We're very good at understanding how to monetize intellectual property in technology. I think that's a common thing we have.
Thank you. Our next question comes from the line of Matt Ramsay with Cowen. Your line is open.
Thank you very much. Good afternoon. I think it would be helpful for either one of you guys to talk a little bit about the M&A philosophy going forward. I think, Tom, you did an excellent job in sort of reiterating the capital return policy. It sort of struck me on a lot of the questions that we got from investors were just sort of a surprise at the size of the CA deal, given some of the prior commentary around maybe focusing on smaller M&A deals. I just kind of open it up and I'd love to hear some philosophy conversation about how you're thinking about maybe verticals or size or any of those things, if there's anything that's off limits going forward, or we should just think about the capital return policy only and nothing's really off the table in terms of M&A.
I'll just open it up at that. Thank you very much.
Good question, Matt. I think that the important message is not much has changed. The business model and what we think drives the returns of the acquisitions that we do is very consistent, and we understand and can appreciate that CA was a bit of a surprise and certainly larger than maybe what many people expected. The reality is it was the right deal for us at the right time as we think about how to grow the earnings base of the company and how to drive value for shareholders over time. If you think back, and this follows up on Harsh's question a minute ago, when we bought LSI we were getting into what you could argue is a very different business. We had a largely mobile business in Avago, RF business that sustains today has grown organically very rapidly.
We were getting into businesses that included read channel SoCs, preamps for hard disk drives. We were getting into enterprise storage and SAS connectivity, delivering to the server market. These were all businesses that we frankly didn't know very well. It was the characteristics of those businesses, just to follow on what Hock discussed, in terms of the intellectual property, the barriers to entry, the sustainability of the businesses, and by focusing on those businesses, that's what allowed us to drive the returns we've seen. We've obviously expanded from there. We've done smaller transactions, but we've also more recently done Brocade
Clearly, that's a systems business. It's largely software. It has an end-user sales force. It's opened our eyes to end customers and what we can do with those end customers. More importantly, it's proven to us that we can manage these businesses. The performance of Brocade over the last year, and we've obviously been quite familiar with the business for more than that time period, has been quite exceptional. CA is really an extension of a strategy that we've been pursuing for a number of years and has driven a tremendous amount of value for shareholders. We look to continue to do that as a way to drive value. Obviously, we're going to deliver on the dividend, which is really important to us.
We have a lot of financial flexibility off the back of significant and substantial operating cash flows to continue to do buybacks and to do accretive M&A. We see opportunities going forward to do that.
Thank you. Our next question will come from the line of Edward Snyder with Charter Equity Research. Your line is open.
Thanks a lot, Hock. Thanks for the explanations. One of the things that struck me, though, in your initial comments when you were talking about your new customer base with CA, because they do open up a lot of clients that you don't have with your semiconductor business. You were talking about porting in some of your networking and compute solution, storage solutions to them. But right now you're selling semiconductors, and a number of your big customers, you mentioned them, Google, Amazon, Microsoft, build their own boxes using your silicon. I would imagine most of CA's customers do not do this. How do you think you could port your semiconductor products to these new customers without getting into boxes? Or are you thinking about that? Or do you think they'll start up such endeavors to start porting through?
I guess I don't understand how your existing products are going to be ported to these new clients without some sort of intermediary or white box guy or something. Maybe I'm confused there, but if you could explain that, I'd appreciate it.
Sure. I think that's a very insightful question you came up with, and you're right 100%. One, we're not interested in going into boxes. You got that right. Or systems. We don't need to do that. We have the key ingredients. We have the chips, the engine. Say, take a box, be it a switch router or any of those things. We have the software. If you want to, some of these end users now, and you're probably aware of that to some extent, some of the big operators now are starting to want to build their own data centers. They've come to us and asked us to enable them to build their own data centers. These guys are very aware how the cloud guys doing. The cloud guys use our own silicon engine, our own merchant silicon, in many cases.
Some of our initial software SDK, in many cases, they even write their own software, and they then go to ODMs. The ODMs in Taiwan, in China, anywhere else, to put a box, to build a system, a box. We help to enable that. Obviously, we have to. That's how the cloud guys do it. There's no reason why an operator like AT&T with Domain 2.0 cannot and is, in fact, executing on that basis, or any other large enterprise who uses, who has a build-out on their own scale, fairly substantial data centers, why they can't even do that on their own. As long as you have the core IP, the core technology, which is the engine, the software, everything else ties together, and there are lots of ODMs out there. We call that white boxes, I believe.
They have a choice of doing that or continuing to buy from their traditional sources. What we are able to do now with our direct access to CA customers is establish strong strategic and strong engagements with those end users, substantial end user enterprises or end users who would want to start doing it themselves in order to not just do it at low economics, but in order to access directly the latest, I call it leading-edge silicon software products, technology, which will enable them to build data centers just as leading edge as what's available in the cloud. We have seen that request coming in, and we are basically responding to it. This is not a pipe dream.
Thank you. We have time for one more question on today's call. Our last question will come from the line of Romit Shah with Nomura Instinet. Your line is open.
Yes, thank you. Thanks for squeezing me in. I definitely appreciate that you guys think very strategically about the deals you do. If I just go back and think about your M&A track record over the last several years, it just has struck me as being financial deals first and foremost. Hock, the playbook has been, at least my impression, has been you'd flash SG&A by a significant amount. You'd cut R&D while basically raising prices at the same time. CA, given the legacy nature of their technology, has a lot of people believing that this business may turn out to not be as sticky if you take the same approach. Could you just comment on that, please? Thank you.
That's a question that's wrong in so many fronts, I don't know where to begin. Let me try. Number one, we have a history of acquisitions and integrating very well. Those are not financial deals. They end up, as I said, as great financial returns. They're not. We operate them under a single umbrella. We operate them, I hate to say it, very well, but we also operate them very focused. When you're focused, as I said before, you don't over-invest in R&D. When you're focused, you don't over-invest in even selling. When you run your company in a business model that is simple, you don't need a huge amount of SG&A, as indicated. It's not about cutting SG&A, it's about simplifying a business process with a portfolio of very strong businesses, where you focus your spending on R&D to enable you to keep being ahead of the pack.
That's really our model. It's a business model that we've executed over the last six years through multiple acquisitions. It's also, as I mentioned in an earlier response, how well we integrate into the model, which ties to being very focused on what you pick as core businesses you want to keep investing in, and divesting. Still keep out of it, you don't have businesses that you do not see as being sustainable. Part of it is sustainable, which lends the term franchise. When we buy companies over the last five, six years, just as many companies and businesses we add in, if you look at the finer print, and we have that data, we divest as many businesses out there.
We just let it go, because those are businesses we believe are not sustainable, are not strong, are commoditized, and where you don't have that advantage for whatever reasons that comes into play. In that sense, you might call it financial, I don't know. I call it very strategic focus on businesses you can win at. That's how we look at even looking towards CA, which is an interesting part of what you say, because believe me, the mainframe business is very alive and well. Investments are still continuing in the mainframe business. To put it in simple terms, online transactions, a lot of them in the largest enterprises in the world, cannot run without mainframe, with hardware or the software tools that drive it. That's basically all I say to that.
Ours is an operating model and a business model, and the financials is what comes out of a very strong, sustainable, and secure business model.
Okay. Thank you everybody for participating in today's earnings call.
Thank you.
That concludes Broadcom's conference call for today. You may now disconnect.