Good day, ladies and gentlemen, and welcome to the Marvell Technology Group Third Quarter Fiscal 2017 Earnings Conference Call. [Operator Instructions] As a reminder, this conference call is being recorded. I would now like to turn the conference call over to John Ahn, Head of Investor Relations. Please go ahead, sir.

Thank you, Abigail, and good afternoon, everyone. Welcome to Marvell Technology Group's Third Quarter of Fiscal 2017 Earnings Call. With me on the call today are Matt Murphy, Marvell's President and CEO; and Jean Hu, Marvell's CFO. We will all be available during the Q&A portion of the call today. If you have not obtained a copy of our current press release, it can be found at our company website under the Investor Relations section at marvell.com. We have also posted a slide deck summarizing our third fiscal quarter 2017 results in the IR section of our website for investors. Additionally, this call is being recorded and will be available for replay from our website until December 17. Please be reminded that today's discussion may contain forward-looking statements. These statements are based on currently available information as of the date of such statements and are subject to risks and uncertainties that could cause our results to differ materially from management's current expectations. To fully understand the risks and uncertainties that may cause results to differ from our expectations and outlook, please refer to today's earnings release, our latest quarterly Form 10-Q and subsequent SEC filings for a detailed description of our business and associated risks. Please be reminded that all of our statements are made as of today and Marvell undertakes no obligation to revise or update publicly any forward-looking statements except as required under applicable law. During our call today, we will make reference to certain non-GAAP financial measures, which exclude the effect of stock-based compensation, amortization of acquired intangible assets, acquisition-related costs, restructuring costs, litigation settlements and certain expenses and benefits that are driven primarily by discrete events that management does not consider to be directly related to our core operating performance. Pursuant to Regulation G, we have provided reconciliations of the non-GAAP financial measures to the most directly comparable GAAP measures in our third fiscal quarter 2017 earnings press release, which has been furnished to the SEC on Form 8-K and is available on our website in the Investor Relations section. Please note that Marvell Technology Group announced the significant restructuring action on November 2 of our fourth fiscal quarter. During the prepared remarks section of this call, Matt and Jean will be providing details of our third quarter results before our restructuring actions. Then Jean will provide comments on our restructuring actions and provide our fourth quarter fiscal 2017 guidance for continued operations. With that, I would like -- now like to turn the call over to Matt Murphy, Marvell's President and CEO.

Great. Thank you, John. Good afternoon, everyone, and thank you for joining us. Before I review Marvell's results from our fiscal third quarter, I want to provide you with some insight into our recent announcements and strategy going forward. As I mentioned on our last call, we completed a comprehensive evaluation of how our R&D resources are allocated across our product development efforts. This included a thorough review of product lines, discussions with customers and feedback from key stakeholders throughout the company. Our assessment confirmed that Marvell's core strengths are in the high-speed movement of data, particularly in the storage, network infrastructure and wireless connectivity markets. As a result of this review, we are focusing Marvell on these strengths and have developed clear strategies to grow in each of these markets. Let me share the details on our strategy for each market starting with storage. Data storage is in Marvell's DNA and we hold leadership positions in both hard disk drive SoCs and solid-state drive controllers. For our HDD business, our strategy is to maintain our leading position in client drives and continue expanding in the cloud and enterprise storage with our higher-performance solutions and strong OEM relationships. The strategy is already yielding design win momentum, and we are gaining traction in the high end of this market. In SSD, our strategy is to extend our leadership in this fast-growing market by leveraging the breadth of our portfolio and continuing to expand our solution offerings. We continue to set the industry standard across a range of interfaces, from SAS and SATA to our leading PCIe-based solutions. Our portfolio includes cost-effective DRAM-less solutions, custom ASICs, mainstream client products with firmware that provides full turnkey solutions and new generations of ASSPs for the rapidly growing cloud and data center markets. We are the first merchant controller provider to offer products built on advanced process nodes, and along with our proven error correction technology, enables higher performance, lower power and smaller footprint, all capabilities that our consumers desire. Innovation is equally important, and we are also leading in the deployment of emerging standards such as NVMe, which is enabling a new generation of high-performance computing systems. Altogether, we are very excited about the profitability and growth potential of our storage business. Network infrastructure is another area of historical strength for the company. We are one of the pioneers in low-power, low-cost, CMOS-based Ethernet transceivers. Today, we have reenergized this business, and our 2.5 and 10-gig Ethernet PHY switches and versatile multi-core ARM SoCs are gaining traction in the marketplace. Expect us to continue with our recently announced 25-gig solutions for the data center, and private cloud markets are also showing promise. We are continuing to invest in this business and moving up the value chain to grow in this market, particularly in products that target the cloud and enterprise data center. Wireless connectivity is a core strength and one that we've grown organically as we've established Marvell as a market leader. Our strategy is to refocus on WiFi technology and invest in high-performance areas. For example, our innovative 802.11ac integrated wireless solutions deliver industry-leading performance with a reduced footprint at a lower cost. As a performance leader, we are shifting from the low end towards markets that value high performance such as enterprise access points, automotive, connected home and gaming. Overall, Marvell's core strengths align with the broader trends emerging in the marketplace. Literally billions of connected devices, not just appliances but also cars, cities and entire industries are moving to the cloud. Virtual reality, artificial intelligence, autonomous cars, smart cities depend on high-bandwidth video and data multiplied by social sharing. This is creating an explosion of digital traffic. Data needs to be moved and stored not just to the cloud but also to the edge for greater accessibility at ever-increasing speeds. Enterprise infrastructure and cloud providers are being forced to innovate at an unprecedented rate just to keep up. By building on our core strengths in storage, wired infrastructure and wireless connectivity and providing differentiated solutions, Marvell is well positioned to transform how data is moved and stored across a range of markets from the consumer to the cloud. Now as part of our R&D review, we looked at our total portfolio from a product-by-product and holistic perspective. We examined where we had differentiated technology and the path to market leadership that would produce profitable growth. We also identified areas where we could improve the return on our current technology investments and which markets don't make sense for us to participate in. Based on this analysis, we announced in November 2 that we will exit or divest non-core businesses, consolidate R&D sites and eliminate approximately 900 positions worldwide. These difficult but necessary actions focus Marvell on areas where we can win and deliver a higher return on our R&D investments. This restructuring is the first phase in our efforts to improve Marvell's performance. Our next phase includes the continued improvement in gross margins and operational efficiency. We've got a number of initiatives in place increase our supply chain efficiency, including consolidating purchasing and other functions to leverage our scale. We are raising the bar on our R&D investments so that new products have clear differentiation and yield a higher return on investment. Our goal is to achieve 58% to 60% gross margin within the second half of our fiscal 2018. With a renewed focus on our core product portfolio and on markets in which we are a leading technology provider, we are confident that with discipline and execution, we will deliver profitable growth. As we achieve greater cash flow and profitability, we are also committed to returning cash to shareholders. With this in mind, I am pleased to announce that of our Board of Directors has authorized a $1 billion stock repurchase plan. As part of this, we intend to return $500 million to shareholders through share repurchases over the next 12 months while also maintaining our current dividend plan. In summary, we have made significant progress over the 4 months that I've been with Marvell and I want to thank the entire team for their efforts. We've assembled a strong management team, refocused our R&D on the company's core strengths, developed strategies that will put Marvell on a path to faster and more profitable growth and implemented changes to make more effective use of our capital. Altogether, I am very confident that these actions will improve our execution as a company, accelerate innovation and yield greater returns for our shareholders. Now I'd like to review the results of our third quarter. We delivered strong financial performance for the third fiscal quarter of 2017. Our revenue came in above the high end of our guidance, growing 4% sequentially to $654 million. This was driven by stronger demand from our storage and networking businesses as well as from our mobile and wireless sales, which came in as expected. We achieved non-GAAP gross margin of 56.7%, which is a significant improvement from 45.9% 1 year ago. We also generated non-GAAP operating income of $115 million and delivered $0.20 in earnings per diluted share, beating the high end of our range. In terms of business highlights storage revenue came in stronger than expected, driven by growth in both HDD and SSD solutions. Our HDD revenue grew as customers saw continued stabilization in demand for client drives. In SSD, Q3 was a record quarter with revenue increasing significantly both quarter-over-quarter and year-over-year. SSD controller sales now represent approximately 20% of our total storage revenue. In the networking market, Q3 sales of Marvell solutions were better than anticipated and grew 20% compared to the third quarter of last year, driven mainly by continued traction of our solutions for the campus and enterprise markets. In the wireless connectivity market, our third quarter revenues came in as expected. This was mainly due to anticipated declines in wireless connectivity solutions related to mobile handset and low-margin modules. Now I'd like to turn it over to our Chief Financial Officer, Jean Hu, to provide more details for Q3 and our outlook for Q4. Go ahead, Jean.

Thank you, Matt, and good afternoon, everyone. First, I will review our third quarter financial results. Then I will discuss our restructuring actions and provide our fourth quarter fiscal year 2017 guidance for continued operations. As a reminder, our Q3 results do not include any impact of our restructuring actions, which we announced on November 2 in our Q4 fiscal 2017. Our revenue in the third quarter was $654 million, which represents a 4% increase from second quarter revenue of $626 million. Our Q3 revenue included $16 million of storage shipments deferred from Q2. Normalizing the $16 million deferred revenue between Q3 and Q2, our storage revenue in Q3 would represent an 8% sequential increase and 21% increase year-over-year, driven by higher HDD and SSD demand. Networking revenue in the third quarter for fiscal 2017 increased 16% sequentially but grew 20% year-over-year. Mobile and wireless revenue declined 11% sequentially as expected. Mobile handset related revenue was $4 million. Our GAAP gross margin for the third quarter was 56.3%. Non-GAAP gross margin was 56.7%, above the high end of our guidance range provided in our Q2 2017 earnings release. This increase was primarily driven by product mix and improvement of operational efficiency. Our GAAP operating expenses for the third quarter were $285 million. Non-GAAP operating expenses were $256 million, lower than our guidance, primarily due to better operating expense management. Non-GAAP R&D expenses in the third quarter were $202 million and non-GAAP SG&A expenses in the third quarter were $54 million. Non-GAAP operating income increased to $115 million from $27 million a year ago. This represents an operating margin of 17.6% versus 4% operating margin a year ago. Our net other income was $5.5 million. Tax provision was $16 million. GAAP net income was $73 million for the third quarter and non-GAAP net income was $105 million. Third quarter GAAP earnings per diluted share was $0.14, and our non-GAAP earnings per diluted share was $0.20 versus $0.06 a year ago. We have made significant progress to improve our financial performance. Turning now to our balance sheet. Our cash and marketable security were $1.65 billion or over $3 per diluted share at the end of the third quarter. Net cash provided from operations in the third quarter was $121 million. During the quarter, we resumed our stock buyback and our original plan. We returned $87 million cash to shareholders, which included $30.7 million in dividends and $56.5 million in stock repurchases. As noted in our press release, our Board of Directors approved a $1 billion share buyback program. This newly authorized program replaced in its entirety the prior $3.25 billion stock purchase program, which had approximately $115 million of repurchase authority remaining. Now I'd like to recap the restructuring actions we announced on November 2, which included 2 initiatives. First, we are discontinuing specific R&D programs, streamlining engineering process and consolidating R&D site for greater efficiencies. These actions will result in eliminating approximately 900 positions worldwide, which is about 17% of our total headcount. We estimate that the revenue associated with these R&D programs was about 7% of our Q3 fiscal '17 revenue. This revenue is primarily under our mobile and wireless as well as other end markets. We do expect this revenue to decline over the next several years but do have a very long tail. We will classify this revenue under other markets in Q4 so we can provide more transparency of our core business in storage networking and wireless connectivity. We also expect a significant reduction in legal and accounting costs. Altogether, these actions are expected to reduce the annual operating expenses by $180 million to $200 million. In addition, we are divesting nonstrategic businesses with approximately $60 million in operating expense and $100 million in revenue based on first half 2017 annualized run rate. These businesses have gross margins of approximately 40% and will be classified as assets held for sale and reported as discontinued operations in Q4. Our guidance for Q4 excluded estimated results of these businesses, which will primarily affect our other end market revenue category. Now I'd like to move on to our guidance of Q4 fiscal 2017 for continued operations. We expect our Q4 net revenue to be approximately $565 million, plus or minus 2%. Excluding the impact of the $60 million deferred revenue in storage, we expect our storage revenue to be down slightly in Q4 compared to Q3 but will grow year-over-year. Networking revenue is expected to be flat sequentially but show double-digit growth year-over-year. Our wireless connectivity revenue is expected to decline sequentially due to normal seasonality. Our GAAP and non-GAAP gross margins are expected to be approximately between 57% to 58%. Our GAAP operating expenses are expected to be in the range of $322 million to $332 million, and non-GAAP operating expenses are expected to be in the range of $225 million to $235 million. GAAP and non-GAAP net other income is expected to be approximately $3 million. And GAAP and non-GAAP tax benefit is expected to be approximately $1 million. GAAP and non-GAAP diluted share count are expected to be approximately 524 million and 532 million shares, respectively. This will result in GAAP income per diluted share in the range of negative $0.01 to positive $0.03 and non-GAAP income per diluted share in the range of $0.17 to $0.21. Looking ahead, we will detail our strategy for profitable growth, including our long-term target financial model and capital allocation strategy at an upcoming Analyst Day in early 2017. We will send out more specifics on this event soon and look forward to seeing many of you in person. With that, I'd like to turn the call back to John to open up for Q&A.

Thank you, Jane. We will now open the call up to your questions. Operator, we'll take the first question, please.

[Operator Instructions] Our first question comes from Craig Ellis of B. Riley.

The first question is on gross margins. So one, congratulations on the significant progress you're making there in other parts of the business. As we transition from the improvement in gross margins to 56.7% and then 57.5%, can you talk about what changes in the business to drive it towards 58% to 60% over the next 3 to 4 quarters? And should we think about that 58% to 60% being a peak margin level for the business? Or is that just an intermediate level with potentially higher margins down the road?

Craig, this is Matt. So yes, we're clearly pleased with the progress the company has made in short order to really address the cost side with respect to the product margin. Additionally, we're benefiting from improved mix with higher sales of our networking and storage. So that's helpful and helping in the near term. As we look out into the longer-term model we're communicating for the second half of 2018, we see further mix improvements and additional benefit that we anticipate in really addressing cost of goods in a very strategic and meaningful way. As far as anything above 60, we're not going there yet. It's a tough crowd. I think we're pretty happy after this period of time to come out and be comfortable guiding in that range, but we do feel confident that the quality of the business that we're winning and that we're delivering is very good and that the road map of new products that we're going to be introducing also has a gross margin profile that supports that type of guidance. So overall, we're pleased with the progress on gross margin.

That's helpful. And then the follow-up question is more of an operational question. That was helpful to get the summary of which businesses you think are growth businesses and which are more maintenance businesses. As the company brings down its operating expenses and R&D significantly in the near term, when do you think you'll have the R&D allocated to your satisfaction in growth year areas versus maintenance areas? And when do you think we'll start to see sustainable revenue growth out of the business as you've gotten the R&D teams in place and as you make whatever tuning adjustments are necessary with field engineering and the marketing program?

Sure. So let me break it in 2 pieces. So the first part of your question was around when do you think we'll have the R&D movements allocated appropriately to the areas where we believe we can grow. And what I'd say on that is, and Jean gave some numbers on it, what we've really done is edit and address the areas of the company where we had R&D applied that we didn't see acceptable returns and meaningful growth prospects. So on the core businesses of storage networking and wireless, we're pretty comfortable with the R&D allocation there. There are some moving pieces within each of those businesses, but overall, we've been investing in those businesses. So there's some fine-tuning going on. But actually, I'm quite comfortable with the road map, the team and the execution we're seeing out of those 3 businesses. That being said, there are minor tweaks we're making. So really what we're doing is addressing the part of the portfolio that is not part of the long-term road map. Go ahead, Jean.

Yes. Just to add to Matt, really if you look at the restructuring action we are taking, largely it's a focus on those non-core areas which have very large R&D but very small revenue and return on investment. And I think we articulated that over the next 4 quarters, we'll reduce those R&D programs to the point which has a run rate of comfortably quarterly $210 million. So that's the objective exiting Q4 fiscal 2018. I think then the question is we're investing in all those core areas Matt talked about, which we're very comfortable and we do believe the revenue growth will come from those market we're focusing on.

Our next question comes from John Pitzer with Crédit Suisse.

Matt, I guess even if I adjust [ph] for the divestitures and look at an apples-to-apples sequential compare, you're guiding gross margins for the quarter with flattish sequential despite the fact that storage is going to be down. And so Matt, I guess help me understand a little bit better as you drive towards this new model on the gross margin line. Do you think all of your products or the majority of your products portfolio, will sit inside of that gross margin range such that the gross margin variability is not going to be as dependent as it historically has been on the storage business? Or how should I think about that? And I guess how are you achieving sort of flat gross margins sequentially despite a mix that seems to be moving against you a little bit in the quarter?

Yes, thank you, John. So on the gross margin question, if you look at our portfolio overall, some of the portfolio we are investing going forward are above our corporate average. There are a few areas that we are investing and we do think there's a greater potential in the future. The gross margin, actually, they are below our corporate average. So we do have the product mix that will continue to be one of the drivers of our gross margin going forward. As Matt talked to earlier, right, we're going to focus on the differentiation of our products going forward. So over longer term, we'll certainly try to get our gross margins higher. And the second question is on the guidance for Q4. You're right, the mix is changing in Q4, but we are making effort to really improve cost of sales and operational efficiency. So we do get benefit from operation efficiency, but on the other side, the revenue compared to Q3 is lower. We do have the overhead assumptions and all those headwind. So in balance, that's what we're guiding, and of course, we'll continue to be disciplined and try to do better.

That's helpful. As my follow-up, Matt, one of the things that was holding you back from reinstating the buyback was just trying to decide what kind of M&A strategy you might follow. I'm wondering if you could just update us on your M&A philosophy now that you've reinstated the buyback. And should we read anything into that philosophy with the reinstatement of the buyback?

Yes, no, thanks for asking about the buyback. So yes, we're pleased to put a sizable authorization out there, which I think signals our confidence in the company and the business. And we felt it was important to get that communicated and authorized and announced at this call and get moving on it. In the context of the buyback, clearly, we've looked at our capacity and our capital structure today and also comprehended the reality that M&A is real in the semiconductor industry today and we have to think about that. And we do have some thoughts around it. We're not ready to communicate those, but clearly, that's a reality that we have to deal with. So I'd say in the context of an industry that's going through the transformation it's going through, it's on our minds. But first and foremost, we're focused on driving Marvell's organic improvements. That's first and foremost. We think we have a good opportunity with the products that we have. And second, we felt it was important to put out an authorization in the commitment to return significant portion of our capital to shareholders while still maintaining dry powder.

Our next question comes from Harsh Kumar from Stephens.

I wanted to follow up on the first question asked. If I heard it correctly, it was 58% to 60% gross margin in the second half fiscal '18. What are some of the levers that still need to happen in a big way for you to get there? And then I had a follow-up as well.

Harsh, thank you. Yes, to get to 58% to 60% in the second half of fiscal '18, there are 2 major drivers. First is we'll continue to refine our product strategy. Just based on the portfolio Matt articulated, we're investing in storage and networking and wireless. We're going to look at areas where we really can get the return back in the investment in those areas. Secondly, I think within the next several quarters, most importantly is we are focusing on our supply chain efficiency, not only just your typical operational streamlining improvements of cost, but also like testing, consolidating test labs, all those things. We do think we have a lot work and we can accomplish. Some of them, as you can imagine, we'll take some time when you want to consolidate testing facilities across all the global locations of Marvell. So those are the very important drivers, the product mix and cost reduction efforts.

And Harsh, this is Matt. Let me just add to that. I think what we found is on the COGS side, as we brought in a very seasoned and experienced operations management team, we've come to a few conclusions. And one is, as Jean mentioned, there's really tremendous opportunity to rationalize our supply chain in the back end by consolidating some of our spend with key partners. If you think of the drivers of our cost, it's obviously in wafers, obviously in the packaging in the back end. Some of that back end is done inside Marvell, but most of it is done with partners. So those levers are being pulled, negotiating with and partnering with the right people to drive our cost structure in a much more positive way. There's also programs we're implementing that are very structured around leveraging our strong product engineering and test engineering teams to drive yield improvement. That's another big lever we can pull. And then finally is fact of the matter is Marvell's always been very, very strong in small, dye-size and optimized design techniques. So some of the new products we have also have that advantage as well. So I'd say it's very much a comprehensive approach we're taking from design to supply chain on the cost side, which has some short-term benefits, which you're even seeing today in our current results. And we think that those efforts are going to continue to yield gross margin leverage over the next 4 quarters.

And for my follow-up, I think one of the key things that has to happen for Marvell to grow is the wins, in for example, enterprise and cloud both had networking and storage. And I think I heard you guys say that you're starting to get some of those. I was wondering, how should we as investors measure that success? Would you have some metrics like number of design wins this year versus last year or anything else you can give us to bite into that data?

Yes, I think what you can expect going forward as we get into the rhythm of these earnings calls is now that we've gotten restructuring announced, buyback, we've gotten through the strategy, we'll begin to give more color on our business, things like design wins, pipeline and where we see the opportunities in a lot more detail. I'm seeing them firsthand. I'm visiting customers, and I'm very encouraged. But as far as the data points, I think those are things that we'd like to call -- we'd like to cover as we make progress as a company and start highlighting some of those things and then ultimately call them out as we see them in our business results.

That's also another good topic for our Analyst Day that's coming up, right? So we'll have -- you'll have the opportunity to meet with a lot of our business unit heads and especially the networking. So you'll get a lot more details about what you're looking for. I think that's the best point for it.

Our next question comes from Chris Rolland with Susquehanna Group.

Perhaps you guys can drill down a little more on the nonstrategic businesses with $100 million in revs. In mobile and wireless, what exactly was this? Was it kind of a stub that remained in kind of cellular investments? Or was it something else? And then what are we talking about in the other bucket as well? And then lastly, how should we think about the potential drag on gross margins from these businesses as well?

Yes, thank you first. So I'll answer your first question on the discontinued business side. Those businesses are currently under the category of other end market. That's the most I can tell you because we're running a sale process. As you can imagine, at this stage, we will not be able to disclose which business we're trying to run the sales process. And on your second question about other category, so why -- look at all the revenue that gets impacted by those R&D programs we discontinued the investment, they are largely under either mobile and wireless or under our other market. Some of them are probably obviously under mobile and wireless. We do have the mobile computing like application process. There's those kinds of business that Marvell had for a long time. So for the instance, application process, they tend to have a very long life. So we're going to start investing for the future generation for product, but we do expect the revenue to last for several years and declining. From a margin perspective, of course, we're going to try to optimize the gross margin of those products when we don't have an investment R&D anymore. And so we do think from the 2 margin dollar perspective, those business will continue to contribute to the operating margin line.

Okay, great. That was helpful. And then pacing on the billion dollars of buyback. So you guys had 500 over the next 12 months. I guess my first question is, should we just kind of expect that to be somewhat linear? Or do you guys have a different strategy around that? And then secondly, should we just assume another 500 the next year? Or is this kind of a multiyear plan in your eyes?

I think we'll try to do the buyback in the most efficient way we can. And so frankly, when we guided our Q4, we did not include any buyback impact in our guidance. I think for modeling purposes, you probably can streamline it for the year. I think we are really focused on the 400 to 500 right now. The board authorized $1 billion, which is a multiyear plan. And right now, we're very focused on the first $500 million.

Our next question comes from Quinn Bolton with Needham & Company.

Jean, I just wanted to make sure I heard you right. When you talked about the discontinued ops, you said it's about $100 million of annual revenue and I think $60 million of annualized expenses. Did you also say it was running roughly at 40% gross margin?

Yes, yes, yes, that's exactly what we said. This business is about $100 million revenue, about 40% gross margin and $60 million operating expense, roughly.

So as that business gets categorized as discontinued ops in the January quarter, you clearly get a little bit of a positive gross margin effect since you take that 40% gross margin revenue stream out of the continuing operations, is that right?

Yes, absolutely. You're right.

Got it. Great. And then sort of a second question. It sounds like most of the products that you're walking away from on the mobile side or looking to categorize as discontinued ops are in the other category. Just within -- what's left then within mobile, it sounds like most of the wireless LAN will be focused on segments where you expect to continue to invest. I just wanted to make sure that -- I know you're stopping the investments in WiFi for mobile handsets, but is there anywhere else that you're stopping the investment in the wireless LAN? Or are you looking to kind of continue that investment in sort of higher-performance WiFi across the remaining segments of that wireless business?

Yes. Quinn, this is Matt. I'll take this one. So yes, the way to think about it is we have this category, mobile and wireless, and then within that is our WiFi business. It's a piece of it, largest piece today. And we are investing there. When we've looked at that business, the handset portion now is pretty de minimis. It's almost down to nothing. When we look at the WiFi market, nonmobile, and you break it into a few categories like enterprise access points, we actually have a leading position there. When you look at automotive, we've got a leadership position there as well, connected home, high-end gaming. So these nonmobile, higher performance, more robust WiFi-enabled applications is a place where Marvell has got a market position. It's got traction. It's got design wins. And we're going to try to build off of that. So that's how we're thinking about the investment portion of the mobile and wireless category.

In Matt, correct me if I'm wrong, but it sounds like most of those WiFi designs will be more access point solutions perhaps with integrated CPUs rather than lower-end sort of client devices, is that right?

That's the way to think about it, yes.

Our next question comes from Timothy Arcuri with Cowen and Company.

We did not hear your question.

Okay, great. My question is -- first one on margins within storage. And really, how much higher SSD gross margins are than HDD gross margins?

Sorry, we don't disclose that information, and I think that the only comment I will say is that our overall storage margin is higher than our corporate average.

Okay, great. And then I guess a question really for Matt. There's some concern that if you did an M&A deal right now, that it could interfere with your ability to basically take cost out of the current businesses. But you obviously have about $2 billion worth of liquidity between your real estate and your cash and what you can borrow against. So the question is, would you be willing to do a deal while you're still rightsizing the cost of the existing business?

Sure. Thanks, Tim. And what I'd say is you're absolutely right. The key focus of the company right now is to execute, execute on the plan that we laid out to the investment community, lay out the plan that we've discussed internally and really drive that. And if we do that, we think we can unlock a lot of value in the company. We have comprehended what you mentioned with respect to our buyback thoughts, but at this point, again, we're laser focused on the internal development. And we've got all hands on deck on executing.

Okay. So I guess the right way just to think about that would be that you want to take the cost out of the existing business before you go out and do a deal, to be clear, right?

Yes, I think all I'll say is this. That's clearly the priority. We have to keep our eyes open. We're not unaware of what's happening in the industry. And we do think that over time, as we execute our plan, Marvell is a phenomenal company. It's a phenomenal platform. We've got a great brand, great infrastructure, great engineering talent and great scale. And so over time, we think we can grow this company organically and inorganically. But first things first.

Our next question comes from Joseph Moore with Morgan Stanley.

This is Vinayak calling in for Joe. I wanted to follow up on your SSD business. Like can you just talk about the growth you're seeing there, your share position as well as the pricing and the competitive profile of the market?

Sure, yes, happy to take the question. So we're very pleased with our performance of our SSD business. As I mentioned in my prepared remarks, it's now becoming a meaningful portion of our overall storage revenue. It's growing much faster than our overall storage revenue, and it's in a market that's growing quite fast. We have a wide range of technology that we offer there. We participate primarily at the higher end of the market, and we're also seeing traction, especially in the last year or so, even in the lower end of the market where really having not only the hardware but the software and the firmware and the full turnkey solution is required. So I couldn't point to any one individual specific growth driver, meaning all of the areas that we're investing in SSD, all the different submarkets are actually all performing quite well. And we're encouraged by the traction we see, especially as large OEMs who are transitioning from hard drive-based systems to SSD-based systems start thinking about tailoring their storage requirements around their own special needs. And when they do that, they can think about using Marvell technology and controllers in conjunction with sourcing their own NAND and coming up with their own solutions for their tailored application. So we really participate, again, in a broad range, from -- all the way from ASICs at the high end for high-end applications down to broad-based solutions for the retail market.

Got it. That's helpful. For a follow-up, can you touch upon what traction you're seeing for Final-Level Cache and MoChi? That's something you guys were pretty vocal about last year. Like any comments on consumer traction there?

Sure. So the way we look at those 2 technologies is they're both technologies that are -- continue to be in the Marvell road map. With respect to MoChi, we're in production today with devices that are based on the MoChi architecture. But I'd say more broadly, the company really took that vision of MoChi as an interface and as a design standard and actually has embraced modular design and modular technology in a number of different ways outside of just the MoChi-specific interface. So we see modular design is important for Marvell, especially as processor cores want to migrate down to advanced node, but I/O really still can be efficiently designed and leveraged at legacy nodes. So that's the technique we are taking advantage of today. FLC is a little farther out, and that's leveraged primarily from our storage group. So both of them are in our new product design pipeline. MoChi is farther along. And we are going to continue to leverage those technologies that have been incubated inside Marvell for our own product benefit.

Our next question comes from Ian Ing with MKM Partners.

First question for Matt. I mean, Marvell, largely an OEM business. But in your past, you also have extensive experience with distribution representatives and other types of channels. So do you foresee perhaps adding other types of channels to help sell Marvell product? I know IoT and connectivity, some competitors use the channel -- distribution channel.

Ian, great question. And it's not something we talk a lot about, but I'll just spend a little time here on it. So you are correct. Marvell, historically, has been very much an OEM direct account, R&D to R&D partnership type of company. And that's great. I mean, the company is actually -- I'm very impressed with the level of relationships that have been developed over the years in Marvell up to very, very senior levels at our customers. So that's an asset we have, and that's great. We do have business that goes through distribution. And you're right, some of it's in connectivity. Actually, a lot of it is in our networking business, in PHYs that fall under the broad market as well as some of our switches. So that revenue today isn't terribly large as a percentage of Marvell, so it hasn't got looked at in a lot of detail. But in an aggregate dollar amount, it's actually not insignificant. And when I look at our distributor network that we have in place today, it really hasn't been actively managed, in my opinion. It's not been a priority. So think of it this way. We sell a lot through distribution to a lot of customers. And our view here is -- and this is in the next sort phase of Marvell now that we've gotten all of the strategy in place and the restructuring and the things I mentioned. Our attention is now turning to how do we take the technology we have in Marvell and sell it much more efficiently and much more broadly. And distribution can be a great mechanism to do that. And I think there's a lot of low-hanging fruit in distribution today to really leverage that revenue we've already got with those partners, maybe think about how to focus within distribution on some key partners and actually motivate them and enable them to sell Marvell products much more competitively. So that's something I think you'll be hearing more from us on. I think it's an opportunity for the company to get incremental sales with technology we've already developed.

And for my follow-up, now that the strategic review is done, do you think there's any opportunities in terms of the process of green lighting new products investments, perhaps revisiting the hurdle rates or ROI and investment requirements?

Yes, you've got my playbook figured out. So that process in the company -- I mean, again, it's a very, very innovative company. And that innovation, that level of creativity in Marvell has created a lot of technology that's very, very valuable to the company. So that's something that I'm very focused on making sure we maintain as we go forward. But that being said, the rigor in the company around having a very robust process to evaluate all the new investments that we're going to make in R&D, all the new projects, what's the true realistic de-risked return on investment for those. Does it meet the hurdle rate that we need to grow the company over time? Is it differentiated, what's the development cycle time, what's the cost, that whole process in the company is something that I think can really be improved upon a lot. It's something I have a lot of passion around. And myself and the whole management team, as we think about Marvell going forward, are going to be spending a lot of time with our engineering groups in really making sure that what we're putting into the pipeline on this, this very precious R&D budget that we have, is being spent wisely. And again, I think that's just -- like you said on the other question with distribution, it's just another opportunity to optimize what's already working in the company and do a much better job at it. And I think we should start seeing dividends paid on that next year and the year after. But to really manage the company at this scale for the long term, we've got to have a very efficient capital allocation process and R&D budgeting process. And we're putting that in place.

Our next question comes from Kevin Cassidy with Stifel.

Maybe my question is along the same lines. But you had mentioned revitalizing your networking, especially Ethernet networking. I wonder if you could talk in a little more detail about that of -- so you mentioned that 25-gig PHY. What are you going to be bringing to the market that isn't there already?

Sure. So you're right. And I think many of you know the history of the company. Marvell, really one of the ways that made its mark in the industry was that at 1-gigabit Ethernet, the company really took the market by storm and gained the leading position. And along with the pioneering in the read channel area and HDD as well as in WiFi, these are kind of the core businesses that grew Marvell to what it was -- what it is today, excuse me. The company did get behind for a number of reasons. And starting in early 2015, there was a leadership change and -- within the company, within the networking business and a real refocusing in that area. We do see strong demand and opportunity for us both to do some refreshing on the existing portfolio, and we've now started to get products out that are leading products and what we call NBASE-T, which would be things like 2.5 gigabit Ethernet, 5-gigabit Ethernet that are addressing new segments of the Ethernet market as well as 10-gig, which even though that standard has been around for some time, actually has not -- did not represent a material portion of the Ethernet shipment today. We have our 10-gigabit solutions as well, and we think we can participate there, especially because of our strong position in switches. And selling our switches and PHYs together as a solution to our customers is a very powerful prospect and one that one of our largest competitors has done really well for many years. So we think there's plenty of markets to participate in. Marvell is very well recognized in this area. And we have a suite of products from 1-gig, 2.5, 5, 10, gearboxes and other interface ICs, the 25 and beyond. And again, coupled with our switches and our ARM-based SoCs, by the way, it's a very nice solution for a wide range of our customers. So we are investing there and we're seeing traction.

Yes. Just to add to what Matt said, right, so if you look at our networking revenue for the past 4 quarters, we had experienced year-over-year growth consistently. So we do see strong momentum from that business not only during the past several quarters but going forward.

Okay. And Matt, you mentioned that -- my follow-up question was around adding the ARMADA or the ARM-based processes to the strategy, too. Maybe if you can expand on that a little more.

Sure. So I think that was a very good decision that was made -- that was done before me. And that product line has actually -- if you look at our networking revenue, and as Jean mentioned, it's been growing pretty nicely. Actually, the fastest-growing piece of it has been the ARMADA family. It's getting traction across a range of applications and range of geographies. It's doing well in products for campus, SMB, even small cell base stations and things like that. And we also see that there are some legacy architectures out there like PowerPC or MIFs that ultimately customers want to see a migration path to ARM, the ARM architecture, which has really, as everyone knows, become the industry standard. And so as that transition happens and people move to ARM, Marvell's got a pretty competitive offering there. So we do see some positive trends in the SSD area. It's still relatively early days, but we are getting revenue growth. And as I mentioned, it's growing faster than our overall networking business. So we're encouraged.

[Operator Instructions] Our next question comes from Mike Burton with Brean Capital.

I'd like to follow up on the SSD comments you just gave, and congrats on a record quarter there. Did I interpret you correctly that the majority of the revenues currently are for enterprise SSDs? And when you look at those newer opportunities on the low end, Matt, are you looking more at client PC as the low-hanging fruit? Or are you looking in mobile flash controllers?

So maybe I'll take the question. So it's a combination of our base revenue between enterprise and client side. And I think when we look at the continued growth of this business, certainly like Matt said, we're going to focus on the high end. We're doing ASIC for a lot of customers. Those are great foundation business. We'll expand it to the client side, not mobile handset. We're more focused on the PC, the things we can add more value.

Okay, understood. And then I know it's early, but with the changing mix of businesses, can you give us a sense of what seasonality looks like in your segments? And any initial thoughts relative to market trends as we contemplate the April quarter in our models?

That's a good question. As you can imagine, the business -- if you look at the last year also, it has changed very significantly, right. We used to have a very large portion of our revenue from the mobile side. So the seasonality has been all over the place, frankly. I would say going forward, if you look at Q4, when we guided Q4, on the storage side, typically Q4, the seasonality, you have 2% to 4% decline sequentially, we actually -- excluding the $16 million deferred revenue shipment, our guidance actually is better than normal seasonality. On the networking side, I think Q4 typically is flattish sequentially. I think right now, we can see networking and storage are quite consistent. The wireless connectivity side, Q4 is very significantly sequentially decline quarter. As you can imagine, a lot of the consumer, mobile gaming, those are the really seasonally down quarter. And going forward, I think it seems our investment are going to focus on storage, networking and wireless connectivity. And over time, those 3 businesses' seasonality will dictate the overall company seasonality. We'll give you more color next quarter when we have a clear cut of our 3 segments, so we can provide more insight.

That does conclude today's question-and-answer session. I'd now like to turn the call back to John Ahn for closing comments.

Okay, great. Thank you, Abigail. I would like to thank everyone for their time today and your continued interest in Marvell. Please note that we'll be presenting at the Credit Suisse TMT conference in Scottsdale, Arizona, a couple of weeks from now, Wednesday, November 30. So we look forward to speaking with you either there or again very soon. Thanks very much.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a great day.