Good day, ladies and gentlemen, and welcome to the Q3 2018 Marvell Technology Group Ltd. earnings conference call. At this time, all participants are in a listen-only mode. Later, we will conduct a question and answer session, and instructions will follow at that time. If anyone should require assistance during the conference, you may press star then zero on your touchtone telephone to speak with an operator. As a reminder, this conference call is being recorded. I would now like to introduce your host for today's conference, Mr. Peter Andrew. Sir, you may begin.
Thank you, and good afternoon, everyone. Welcome to Marvell's third quarter of fiscal year 2018 earnings call. Joining me on the call today is Marvell's President and CEO, Matt Murphy, and CFO, Jean Hu. Before I turn the call over to Matt, I wanted to remind everyone that certain comments today may include forward-looking statements which are subject to significant risks and uncertainties, and which could cause our actual results to differ materially from management's current expectations. Please review the cautionary statements and risk factors contained in our earnings press release, which we filed with the SEC today and posted on our website, as well as our most recent 10-K and 10-Q filings. We do not intend to update our forward-looking statements. During our call today, we will make reference to certain non-GAAP financial measures.
A reconciliation between our GAAP and non-GAAP financial measures is available on our website in the investor relations section. With that, let me now turn the call over to Marvell's President and CEO, Matt Murphy.
Good afternoon, everyone, and thank you for joining us today. It has been a very busy past eight days for Marvell. Last Monday, we signed a definitive agreement to combine with Cavium, uniting two great companies that together will deliver unmatched innovation and end-to-end communications infrastructure solutions. Today we're here to review Marvell's strong third quarter performance, which was the result of improved execution across the company and growth in our core businesses of storage, networking, and connectivity. From a financial perspective, revenue in the quarter was $616 million, above the midpoint of our guidance. Our core businesses grew on a sequential basis and have together grown 7% year to date. This exceeds the target we set in March during our investor day.
Our non-GAAP gross margin grew 400 basis points from a year ago to 61.6% today, as our world-class engineering and improved execution continued to deliver tremendous value. Operating expenses on a non-GAAP basis came in better than expected, below the low end of our target range we set last November. Result, non-GAAP operating margin grew nearly 900 basis points from a year ago to 28.4%. This demonstrates the power of our business model, and we are making solid progress towards our target of 30%. Finally, free cash flow grew nearly 90% from last year to $194 million. Over the past five quarters, we have returned $860 million to shareholders, more than 140% of free cash flow.
On the operational front last quarter, Tom Lagatta, EVP of sales and marketing, brought together more than 400 sales professionals, field application engineers, and partners from around the world for Marvell's first ever sales conference. This featured a week of intense product and sales training and laid the groundwork for the year ahead. Since joining the company less than a year ago, Tom has transformed our sales management, go-to-market strategies, compensation programs, and customer support. Customers, partners, and employees alike say they are experiencing a new and renewed Marvell, one that is beginning to live up to its true potential. Now turning to our business performance. I'm pleased to report that in Q3, our storage business performed slightly above expectations, growing roughly 1% sequentially. Year to date, our storage business has grown 10%, driven primarily by the strong performance of our SSD portfolio.
We continue to ramp our new NVMe-based solutions, and we are well positioned as the SSD market transitions to this high-performance interface. Overall, we see continued traction of our SSD products, particularly in the enterprise and cloud data center, as our customers adopt our industry-leading solutions. Our HDD revenue came in higher than originally expected, fueled by the ramp of our one terabyte per platter HDD controller solutions. These platforms enable many more of our customers to offer higher capacity and lower power solutions to their customers. The transition to one terabyte solutions per platter is progressing well and highlights the value that Marvell solutions offer customers. The bottom line is that as overall demand for storage grows and as capacity migrates towards the enterprise and data center, we remain well positioned in the storage market across both HDD and SSD. Moving on to networking.
Our networking business was up 2% sequentially and 3% year-over-year. Our success in Q3 was driven by our refreshed portfolio of switches, PHYs, and embedded processors for the enterprise market. This growth more than offsets declines in our legacy portfolio. We are very pleased by the performance of our new products, which target the enterprise campus and data center. Marvell is well positioned to capitalize as end customers upgrade their switching platforms to deliver multi-gigabit, enhanced security, traffic analytics, and greater support for mobility. We also continue to see strong customer interest in our secure automotive Ethernet products. While it's still early, we have established active engagements with the leading automotive OEM and tier 1 suppliers. This technology is a foundational building block for the next generation high-speed in-car networks that are vital for driver assist and autonomous driving.
This is a trend we not only enable but have the potential to accelerate. Overall, we remain excited about the position of our networking business and its traction in the enterprise campus and automotive markets. As we look toward our next fiscal year, we expect continued growth in our networking business on a year-over-year basis. Moving to connectivity. This business enjoyed a strong quarter, driven by strength in automotive, enterprise access, and smart voice-connected devices. We continue to shift our strategy towards this market's high-performance segments, focusing on applications that value the innovation that Marvell delivers. In the coming weeks, we will announce our next generation 11ax portfolio, which will deliver best-in-class features and performance for enterprise access points, carrier gateways, set-top box, and automotive applications. Today, Wi-Fi networks are supporting more users and more devices than ever before.
Marvell's implementation of the new 11ax standard will bring our customers four times greater capacity, support for the greatest number of users, symmetrical uplink and downlink performance, and greater coverage in all deployments. We started sampling these products more than six months ago and are now starting to see strong customer traction. In summary, our strong performance in the third quarter highlights our momentum in transforming Marvell into a leader in providing higher-end infrastructure solutions in storage, networking, and connectivity. Looking forward, we anticipate that storage and networking will represent a growing percentage of our business and expect to see continued expansion of our gross margins in Q4 and continuing on into FY 2019. Let me close by saying Marvell has made outstanding improvements as a company, and I'm very proud of our team's performance.
I want to thank Marvell's employees around the world for their part in making our success possible and look forward to even greater success in the future. Now let me turn the call over to Jean.
Thanks, Matt, and good afternoon, everyone. I'll discuss the highlights for our third quarter of fiscal 2018 and provide our current outlook for the fourth quarter of fiscal 2018. Revenue in the third quarter was $616 million, about the midpoint of the guidance provided in August. Our core business of storage, networking, and connectivity grew 1% year-over-year and were up 7% year-to-date. Storage accounted for 51% of revenue, grew 1% sequentially, and declined 4% year-over-year. As Matt stated earlier, our storage fiscal 2018 year-to-date revenue grew 10%, driven by the ramp of our SSD business and increased presence in the cloud and enterprise segment with our HDD and SSD solutions. Networking accounted for 24% of revenue and grew 3% year-over-year. Connectivity had a solid quarter and accounted for 17% of revenue and grew 19% year-over-year.
Finally, other products accounted for 8% of revenue and declined 22% year-over-year, consistent with our expectations. GAAP gross margin for the third quarter was 61.3%, and the non-GAAP gross margin was 61.6%, a record level for Marvell, and an increase of four percentage points from last year. GAAP operating expenses were $228 million, and non-GAAP operating expenses were $205 million, both below the guidance range we provided in August. These results mark an important milestone in our restructuring effort as we beat our operating expense reduction target set out a year ago. GAAP operating margin was 24.3%. Non-GAAP operating margin was 28.4%, versus 19.5% a year ago, highlighting the strength of our financial model and continued progress toward our target non-GAAP operating margin of approximately 30%.
GAAP earnings per diluted share was $0.30, and the non-GAAP earnings per diluted share was $0.34, an increase of 62% year-over-year. Let's now turn to our balance sheet. At the end of the third quarter, our cash and marketable security were $1.7 billion, or roughly $3.40 per non-GAAP diluted share. Our free cash flow generation was $194 million. Moving to returning capital to shareholders. In the quarter, we returned approximately $170 million to shareholders, $140 million in share repurchase, and $30 million in dividend. Now turning to our fiscal fourth quarter of 2018 guidance. As a reminder, this is a 14-week quarter. We expect our total revenue to be in the range of $595 million-$625 million. At the midpoint of our guidance, we expect our storage revenue to be approximately flat sequentially, in line with seasonal trends.
We expect our networking revenue to continue its growth momentum in the fourth quarter and grow in the high single digits sequentially and the double digits year-over-year as new platforms with our refreshed enterprise access solutions ramp into production with our leading OEMs in both China and North America. We expect our connectivity revenue to decline in the mid-teens sequentially in Q4, largely in line with the seasonal trend. We expect other revenue to be flattish sequentially. We expect our GAAP and the non-GAAP gross margin to be approximately 62%. We expect our GAAP operating expense to be in the range of $240 million-$246 million, and the non-GAAP operating expense to be in the range of $215 million-$220 million.
We anticipate GAAP income per diluted share in the range of $0.23-$0.29, and the non-GAAP income per diluted share in the range of $0.29-$0.33. With that, we'll now open the line to Q&A. Operator, we'll take the first question, please.
Thank you. Ladies and gentlemen, if you have a question at this time, please press star and then one on your touch-tone telephone. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. To prevent any background noise, we do ask that please place your line on mute once your question has been stated. Our first question comes from the line of John Pitzer with Credit Suisse. Your line is now open.
Hi, this is Charles Kazarian on behalf of John Pitzer, thanks for letting me ask a question today. I'm looking at your cash return. You've now bought back about, I believe, $700 million of shares over, call it, the last five quarters or so and reduced your shares by about 6%. Given the recently announced acquisition, could you perhaps speak to kind of your expectations for capital return, i.e., how should we think about potential slowdowns in buybacks around the timing of M&A? Thank you very much.
Yeah, this is Jean. We have stopped share repurchase in conjunction with the current acquisition announcement because, as many of you know, in conjunction with this transaction, we're actually going to issue shares. We're not allowed to do buyback until the closing. We are continuing with our dividend payment each quarter. Our plan, after we close the transaction, certainly is restart the buyback initially maybe in a slower pace. Right when we get our debt-to-EBITDA ratio to our target range, we'll continue to commit to return 50% of free cash flow to shareholders.
I guess just possibly as a quick follow-up. The Auto Ethernet business, I believe, was excluded in kind of your long-term revenue growth guidance at your Analyst Day. It seems like your commentary is kind of becoming incrementally more positive on kind of progress. Can you kind of just talk about the trends you've seen through this year and how comfortable, at what point you might become comfortable around kind of including that as a part of your long-term revenue guidance? Thank you.
Yeah, great question, and you're right. When we gave our long-term model at our first Analyst Day we did earlier this year, we intentionally didn't include automotive such that we wanted to give a kind of a very clear picture of the current business and portfolio of Marvell Technology. Since that time, over the last six months, we've continued to make great progress in the automotive area. We've hired a general manager, we've assigned a dedicated team to this effort, and we've made great progress with customers and our traction. Felt it was appropriate to at least start giving some color around our progress there.
We do think that this revenue is going to be very modest in its ramp, and I would say you should expect us to start including that as a model update sometime next year when we probably do our next Analyst Day, would be sort of how I would think about that from a long-term modeling point of view. For the investors out there, I think as we sort of make progress towards that date, we'll give color as it's meaningful to people.
Thank you. Our next question comes from the line of Blayne Curtis with Barclays. Your line is now open.
Hey, guys. Thanks for taking my question. Nice quarter. Matt, maybe you can just talk about in the storage business, SSDs had a huge ramp this year. I think you talked about maybe 25%-30% of that bucket in the second half. Just curious if that's on track. Can you just talk about your visibility into next year? You mentioned enterprise. Maybe if you can just thoughts on the size of that market and your positioning there.
Sure. Yeah. Hey, Blayne. A couple of things. One is I'd say, we've been pleased with the ramp and it certainly continued, and what I would say is we're still on track to get towards the, I'd say the high end of the range that we mentioned earlier in terms of the 25%-30% contribution from SSD in the fourth quarter. Again, I think we're pleased with that performance and that trajectory. As far as the enterprise side and also the solutions that are going into the cloud, that continues to become a more and more meaningful piece of our business. I don't have the exact breakout here.
I don't know if, Jean, you want to comment, but that continues to be more and more important to us, and I think from a visibility point of view, we only see that continuing in calendar or in fiscal 2019. I'd say that plus increased growth in the enterprise and cloud HDD drives as well, I think will also be a growth driver. I think between the two, SSD and HDD, from an end market pull perspective, we see growth there, which is one of the reasons why we're continuing to be excited about our storage business In the near term, as well as through the next year in terms of the growth potential that exists.
Thank you. Once again, ladies and gentlemen, if you'd like to ask a question at this time, please press star and then one. Also, we do ask that you keep your questions to one question at this time to be sure that we can answer all questions today. Our next question comes from the line of Karl Ackerman with Cowen and Company. Your line is now open.
Hi, good afternoon, everyone. Congrats on a great quarter and guide. Matt, I wanted to focus on the networking business. Based on your results and outlook, it appears you have a much stronger than anticipated traction within your networking business from the 25 new products you recently introduced. How should investors think about the ramp-up of that business over the next four quarters, particularly given some of these design wins are now just starting to see traction this quarter? I mean, why wouldn't the year-over-year growth you're seeing today continue for all of calendar 2018? I have a follow-up, please.
Okay. Yeah. No, great, Karl. I'd say, we're very pleased with how the networking business evolved throughout the year. I think there was a crossover quarter where this legacy business was declining, and our message to investors was, "Hey, just wait till the back half," because we did have some visibility into ramps of these new products that you mentioned, and it's actually happening. I think we're very pleased with sort of the outlook we have for the fourth quarter. That clearly has contributed to the strength of the guide in terms of it relative to consensus. That's all positive in tracking. It's about 10% year-over-year. If you just look at maybe a little bit higher, I think 11% year-over-year is sort of the implied, if you just take the midpoint of our guidance.
We do think that that growth will continue because it is driven by new product ramps and new programs by our customers, and we do think that that's going to continue through fiscal 2019. I can't model it for you on a quarterly basis, we do think that that growth rate that we're on, we should be able to continue if you just think about a year-over-year perspective.
That's very helpful. If I may, during your Cavium call about a week ago, there were several questions on your appetite for M&A beyond Cavium. I think what was missed is your inclination toward further pruning of your overall portfolio. Specifically, is your connectivity business or any of your other businesses sacrosanct in a sense of, we're acknowledging that competition and connectivity is fierce, we have this 802.11ax wave that should be significant in a semi-autonomous vehicle and connected home world. Your thoughts there will be appreciated. Thank you.
Sure. Yeah. No, I'm glad we'll pivot to the portfolio question, not the M&A question, because as everybody knows, we've just announced eight days ago we're entering into a transaction. To talk about the next one is way too premature. From a portfolio point of view, what we said on the call was with respect to Cavium, that there was no planned divestitures of that business. On our own business, we continue to look at that very pragmatically, and we'll do the same, by the way. You should expect this from this management team, that we'll always take a very full view of the entire portfolio of all of our businesses on a regular basis to make sure that we're tracking. With respect to connectivity, you're correct. There's actually a pretty exciting new wave of capability that you're going to get with AX.
We, by design, have been a little bit reserved in sort of our promotion of this. We sort of prefer at Marvell to make progress with customers, get the products to where they're much further along before we start talking about them. As I indicated, you will see some announcements probably before the end of the year here on what our portfolio will look like. We do think that certainly that wave for AX does play into the end market focus we have, whether it's upgrading of enterprise access points in corporate campuses or whether it's, as you mentioned, in semi-autonomous vehicles. Connected home clearly has got tremendous bandwidth requirements that are going to be enabled by AX. I think there's a number of applications there. As you mentioned, connectivity is a more challenging business because it is standards-driven.
We're hopeful that with our portfolio repositioning of our current connectivity business to the higher performance end, we think that's the right way to go, and AX certainly plays into it.
Thank you. Our next question comes from the line of Vivek Arya with Bank of America Merrill Lynch. Your line is now open.
Thanks for taking my question, and congratulations on the good execution. Just a quick clarification and the question, maybe for Jean. Jean, how should we think about April in terms of seasonality on revenue and cost, given the extra week in January? Then maybe, Matt, just to push you a little bit more on the connectivity side, how do we gauge your consumer versus enterprise mix within connectivity? Because I can understand the rationale for being engaged with enterprise, because that's what Cavium also strengthens. What is the rationale for staying engaged on the consumer side, then, connectivity? Thank you.
The question, we guide one quarter at a time. On April quarter, maybe I can give you some color on the operating expense side. Q4 is the quarter we have 14 weeks, but Q3, we do go back to the normal 13 weeks. However, as you know, that's the quarter the payroll taxes start to come back. I would say, on balance, because it's a 13-week quarter, but you add $3 million to $5 million of payroll tax. It's probably from OpEx modeling perspective, it's largely flattish with the Q4 OpEx. That's the color I can give it to you.
Okay, Vivek, let me drill down a little bit into connectivity to try to answer your question more fully. Just as a backdrop, maybe to sort of take you from where we were to where we are. When I came into the company and we all got involved, the connectivity business was very challenged in the company. The mix, even if you go back like 18 months ago, was a lot more consumer-oriented. Gross margins were dramatically lower and was not a business that I think looked like the model that we were developing. If you fast-forward to now, I think we've made tremendous progress with that product line. We've first of all improved the overall cost structure of the company, across every business we have, that's contributed to improvement in margins.
As we said, maybe three or four calls ago, we exited some low-margin module business that we had and some lower-end segments. I mean, things like, we were in toys as an example, which really was not where the profitability was. This journey we've been on to focus not just, by the way, on enterprise access points, but automotive has done well for us and also high-end consumer, which is, you could say that that's high-end connected home or value-added high-end consumer devices where we can add value. That is part of the portfolio, and that is part of the portfolio that we're going to support going forward. We've clearly elected to not go as broad as we were and chase some of the real low-end applications that I just didn't think added value.
From a go-forward standpoint, we're more focused on delivering value and managing around profitability in that business and not necessarily driving top line as hard, but making sure that we add value there. You should expect, heading into next year, we'll still have a mix that's got some consumer in it. I would say it's high-end consumer, but ideally, it's going to see a continued growing portion of more enterprise, automotive, and high-end connected home applications. I think that mix will be more favorable to us, but we're not going to chase growth at the expense of margin. Hope that's helpful.
Thank you. Our next question comes from the line of Quinn Bolton with Needham & Company. Your line is now open.
Hi, let me add my congratulations on the nice results and guidance. Matt, just wanted to come back. Last quarter, you talked about seeing some headwinds in terms of the inventory in the HDD market as some of your customers were going through factory transitions. Is that mostly wound up, or do you still see some inventory in the channel that's impacting the HDD segment?
Yeah. Thanks, Quinn. As you point out, for different reasons, there were inventory challenges over the last few quarters in the HDD customer base. It's hard to know precisely on when it's going to burn and normalize, but the readout from the team and from the customers who we have pretty close collaboration with is that we believe by the end of this quarter that inventory levels in aggregate for the HDD customers will normalize to where they want to run them. Within that mix, there'll probably be some where they want to keep a little bit more inventory and some of them where they're a little bit leaner. On the whole, the high-level readout is based on our current projections, we think that inventories exiting the quarter will be in line and rebalanced to where our customers want them.
If that's the case, then we can sort of go from there.
Great. Then if I could just squeeze a quick question on the networking. You talked about some nice design win traction for the new Ethernet switch in PHY. Just curious, in the enterprise, what you're seeing in terms of adoption for 2.5 and 5 gig Ethernet, is that starting to gain traction in terms of the design wins, or is pricing still keeping the enterprise market mostly at 1 gigabit? Thanks.
Sure, yeah. I'd say the near-term strength is not from multi-gig, although there's just a little bit of it. We see that as more of a fiscal 2019 driver. Actually, the wins that we're seeing and the ramps we're seeing are either on platforms that our customers have where they've adopted Marvell, so some of those are happening from, by the way, both in our Chinese customers as well as North America customers, so it's multi-customer, multi-geo. It's not multi-gig. It's either new higher density gigabit Ethernet switches and PHYs or some of our 10-gig products. We see those continuing through fiscal 2019 as well as multi-gig starting to ramp as well.
That, again, kind of going back to the question I believe Karl had, those are the sort of some of the trends that we see that lead us to feel very good about our networking business and this kind of refreshed effort from the customer base to introduce new campus and enterprise boxes into the market where we're attached to those. We've been thinking this is going to happen for some time, and it looks like it is.
Thank you. Our next question comes from the line of Craig Ellis with B. Riley, your line is now open.
Thank you for taking the question, congratulations on the very good execution and the nice outlook. Matt, I wanted to just follow up on the comments in your prepared remarks about the rebuilt sales team under Tom Lagatta. Since so much of what's happened at the company over the last 12 to 18 months has been significant operational changes that have then unleashed very significant inflections in different parts of the business financially. As we look at that rebuilt sales team, where should we expect to see the greatest impact in the income statement next year? Is it with an acceleration in growth in some of the segments, or is it something different with gross margins, given how they may or may not be comped differently versus the prior sales team?
Sure, yeah. Thanks, Craig. Yeah, I think I definitely called it out in my prepared remarks because we do think as we head out of this quarter that we just announced the results for, is sort of the quarter in which we've finished the restructuring actions, now we're looking at growth. So that pivot and the sales conference we had, which again, I also called out because interestingly enough for a company of this size to have never had a worldwide sales conference, I think was something that probably looking back could have helped. This was a tremendous effort to get the entire team together. Coming out of that, I think that our mantra and our message to these guys was from Jean and I and the whole management team was really, look, it's time to pivot from this mode we've been in to growth mode.
That means driving higher revenue growth. Now, we don't have a gross margin bogey on the sales team. From my point of view, that's the business unit's responsibility. They're out there chartered with pretty aggressive design win targets and goals. Given the dynamics and the customer base and the reception of this sort of renewed Marvell, I would say also incrementally, this is still pretty new, but the outpouring of customer support from the announcement of the Cavium acquisition, that combination has been viewed extremely positively by our existing customers who just like the fact that we're going to be bigger, more viable, have a broader product line as well as the Cavium side. On the whole, Craig, just to bring it back, it's really about driving incremental revenue growth in the top line, and it's really driven through design wins.
The product set that we have, it's not a catalog product line. Typically, those are new wins on new platforms that we drive. Having a very competent sales force with great relationships and great management can move the needle. We've got some stretch goals on them, we're hopeful that they'll start delivering in the next year and certainly beyond that, we can start expecting to see some impact.
Thank you. Our next question comes from the line of Gary Mobley with Benchmark. Your line is now open.
Hi, everyone. Thanks for taking my question. Jean, I think you've mentioned in the past an expectation that the other revenue, which is primarily printer engines, will decline over time. Based on your guidance for the fourth quarter, that business should trend basically flat sequentially throughout the year. How should we think about the long-term expectation or rate of decline for that other category? Then just one follow-up relating to the debt covenants. Now that you've had eight days to finalize the financing for the Cavium acquisition, can you tell me specifically whether or not there's any covenants precluding Marvell from making additional acquisitions?
Okay. The first question on the other product line. First, our other product line is primarily comprised of customer ASIC and the printer business. In the past, during our Analyst Day, we guided for fiscal 2018, this business will decline about more than 25%, which largely right now, if you look at it, is consistent. It's probably between 20%-25%. Going forward, we actually said during our Investor Day that our printer business actually is quite stable going forward. It's slightly declined but relatively stable. We actually see fiscal 2019 and beyond, this other category is going to decline more like 10%-15% year-over-year. Q4 certainly is a little bit unusual. It's actually flat, but it's consistent with some of the printer businesses tend to be stable quarter-over-quarter and going forward.
Hopefully that will help you to model other category going forward. The second question on the debt side, it's really early. We literally just closed the transaction and with the Thanksgiving holiday, we are not negotiating the debt covenant yet, but we'll update you when we get there. This is too early.
Sandra?
Thank you. Our next question comes from the line of Kevin Cassidy with Stifel. Your line is now open.
Thanks for taking my question, and congratulations on the great results. Your SSD business is doing very well. Can you comment at all on what you're seeing as far as availability of the NAND flash devices, just for the broad market?
Sure. Hey, Kevin. Yeah. That's an interesting one, and we monitor that as closely as we can through our partners. I think there's different points of view, but I'd say in general, since the last time we updated everybody, we see that supply loosening up a little bit earlier than I think sort of where people thought. This thing's kind of been sliding every quarter, but there was a view at one point that this wouldn't be until second half of 2018, people were saying. It seems like now our read-through is probably in the beginning part of 2018 is when supply starts to loosen. I just say that very qualitatively, that's a mix of inputs, but it seems like in the limit, it's probably moved up from its sort of most bearish outlook of when supply was going to free up.
We obviously think that's a good thing for Marvell. Although our current customer base is generally speaking the folks that control demand, we do have a number of our customers where we're designed in, where they've been supply limited. I think it's just good for the overall market, us in the industry, if supply loosens and demand can make progress because of our competitive positioning there. Anyway, I think it's probably early 2018, but we'll certainly keep updating folks as we hear more.
Thank you. Our next question comes from the line of Joseph Moore with Morgan Stanley. Your line is now open.
Great. Thank you. I wanted to ask about the 14-week quarter in the context of revenue. I think I heard you talk about the OpEx implications, but is there any revenue ramifications of that extra week as we think about January and April?
From the revenue perspective, when you look at the 14-week, we actually have the whole holiday during December, as a company, we actually have a holiday shutdown for that week. When we look at the revenue side, this guidance, it's actually a balance when we look at the 14-week, look at the holiday week. Overall, seasonally, it's quite consistent from what we see. Of course, once the quarter progresses, we'll see if there's additional information we can incorporate into our view of the revenue. Right now, that's how we see the quarter.
Okay, great. Thank you. Then since you mentioned 11ax, I wonder if you could just talk about the timing of actual deployment and what are the gating factors. Do we need to see 11ax devices before we see infrastructure? Is infrastructure lead the way? Is there still standards definition that has to take place before we can see that roll out more broadly?
Sure. Yeah. Thanks, Joe. Just on the first point, I would say it's kind of interesting on your 14-week quarter question, because for sure the OpEx is there. You're paying people for 14 weeks, you get into this revenue thing, and as Jean mentioned, it's a little tricky. Anyway, on AX, I think it is quite fluid still. That's a little bit of why we've held back in sort of all these proclamations because there is some of the standards that still need to come together, which is, I think, why waiting probably has helped us. As far as how it's going to roll out, I don't have a real clear view yet of sort of what's the chicken and egg of when you see devices versus the access side versus the client side. That's something certainly I can get back to you on.
I just don't have a good feel for it intuitively. Maybe Jean, if you have a view. Otherwise, Joe, something that I can certainly circle back with you on after I talk to my guys.
Thank you. Our next question comes from the line of Ross Seymore with Deutsche Bank. Your line is now open.
Thanks for letting me ask a question. Matt, I wanted to focus on the storage side of things. You started the year with really strong growth year-over-year, you had some tougher comps now, it seems like it's going to re-accelerate, not just in the fourth quarter, but next year. Can you just talk about the moving parts and what sort of growth rate do you think is sustainable in that business as the SSD and the enterprise side appear to be rapidly getting bigger than the traditional client side of the HDD business?
Yeah, sure, Ross. You're right, we started the year strong and then I think a couple things went on. The one was the inventory situation, which resulted from all kinds of different factors, which I think we belabored in multiple calls. There's been a little bit of a slowdown on the HDD side year-over-year, obviously offset by storage, which has really had tremendous growth. It's interesting because when we look at it from our point of view, calendar 2018, maybe even early 2019, but just sort of think of it in the next year or so, the HDD and SSD controller SAM basically converges, which is interesting. At some point, soon, the SSD opportunity will be as big as the HDD opportunity. Market's clearly developing. We've been growing at a rate that's faster than that.
When we look at the moving pieces, certainly, I think through a scenario in which inventory is now normalized and we head into 2019, I think that on the HDD side, we have several tailwinds going on. The first is we have this transition, as I called out, from 500 gigabytes per platter to one terabyte per platter. That gives us some incremental ASP uplift when that transition happens. Second is that as we continue to make progress in the enterprise and nearline side, those designs have some incremental growth left in them. We also have preamplifiers, which we didn't talk about in the prepared remarks. Again, that's a smaller business, but we do have design wins that are going to be ramping next year.
We haven't sized those yet, but I think you should expect in future calls as that revenue starts to come in, we would be able to size it. It sort of paints a picture where ideally we stick to our goal of trying to manage the HDD portion with all of those moving pieces to a flattish level, if possible. When you layer on top SSD growth, we do continue to see SSD as being a kind of mid to high single-digit grower if we're successful. Now, that's not been generally the market view and consensus that we can do that, but that's been our plan.
I would say, just final point is incrementally since we had our Analyst Day and we called out that growth rate of that SAM and what our opportunity was, if you go sort of now we're six, eight months beyond that, we feel incrementally better that our revenue outlook is probably a bit higher than we thought on the storage side. Again, we like our prospects for that reason. We'll have to see how the next year plays out. We're not doing our fiscal 2019 year plan, but just to give you some qualitative updates as we sort of bridge from the end of the year here and the outlook, those are some of the things that you might want to think about in your model and how you sort of measure us going forward.
Yeah, to just add to what Matt said, right? If we base down the middle point of our guidance for Q4, our fiscal 2018 storage revenue actually will grow like 8% year-over-year. It's better than what we said during our Analyst Day.
Thank you. Our next question comes from the line of Harlan Sur with J.P. Morgan. Your line is now open.
Good afternoon, congratulations on the solid quarterly execution and margin expansion. If I look at the SSD market, enterprise is the fastest-growing segment that's roughly about a 15% figure if I look out over the next kind of three years. Here, you guys have been locking up a pretty solid lineup of customers in this segment. I think you guys have captured Western Digital, both the WD and the SanDisk product lines, also Seagate. My sense is that enterprise SSD, as a % of your total SSD business, is now a pretty big part of the mix. Is it approaching 30%-40% of your total SSD revenues? Then if you can just help us understand where you're winning here. Is it more enterprise or more cloud-based SSDs? Thank you.
Okay, thank you. I think your question on the enterprise of SSD portion, yeah, you're in the ballpark. It's actually growing significantly, so I would say the change has been really rapid, but you are in the ballpark. Going forward, certainly we continue to see the revenue ramp overall SSD business and also specifically on the enterprise side. We do expect it's going to be an increasing larger % of our SSD revenue going forward.
Great. Then if I could just to follow up, how much is the team benefiting from Intel's Purley server CPU launch, which commenced in the September quarter? I know that you guys have a good partnership with Intel on both 10 gig and 25 gig Ethernet. Did this contribute to the growth in the networking business in the third quarter, and will it be contributing to the growth here in the fourth quarter?
It certainly helps, to be really frank, that's a very small portion of our business. As we said in the past, our networking business is more focused on the access side, enterprise access side. The overall data center exposure is relatively small, but it does, quarter-over-quarter, if it was a very lower base, it does help a little bit.
Thank you. Our next question comes from the line of Mark Delaney with Goldman Sachs. Your line is now open.
Yes, good afternoon. Thanks for taking the questions. Just a couple follow-up questions from me on some of the prior topics. The first one is to follow up on connectivity. Matt, I understand your increased focus on higher performance areas within connectivity. I was curious if you could talk specifically around game consoles. I may have missed it, but I didn't hear you talk about that, and I'm curious if that factors into any of the areas that Marvell may de-emphasize. So maybe you could help us quantify how much exposure there may be to that product area. Sure. Yeah. As I said, I think there's what we call the higher performance segments, then there's segments that are more challenging. Gaming's probably somewhere in the middle, depending on the application.
We're not gonna get into individual kind of sub-parts of connectivity and their trajectory and momentum, because at some point it just becomes too small, especially when we look at connectivity and we pro forma that business with respect to when we end up combining with Cavium, it becomes a portion of our business where I think to call out individual pieces on their direction is probably not big enough. That's one where, in that segment and other consumer segments, we just need to stay agile and figure out what makes the most sense for us from a profitability point of view. Again, can we add value? Do our customers want us in there or not? If we can't add value, then we won't participate because that just may not be our business. That applies across a number of connected devices that are out there.
Some care and some don't about the technology we bring to the table.
Thank you. Our next question comes to the line of Srini Pajjuri with Macquarie Capital. Your line is now open.
Thank you. Good afternoon, guys. A question on the SSD side, Matt. Obviously, the units are growing nicely there, you said you're pretty optimistic about next year as well. I'm just curious as to where we are in the PCIe transition within client as well as in enterprise. As we go through the transition, do you anticipate your ASPs to also benefit? Thank you.
Sure. I'll give my comments and maybe Jean wants to add, too. In general, when you look at our revenue today, PCIe or NVMe solutions are still pretty small as a percentage of the total. Now, again, our total is getting bigger, don't want to discount the fact that we're shipping real revenue. Between SATA and SAS, and by the way, some of our SAS solutions also enable NVMe, the bulk of the revenue today is not in that area. We do see that progressing into next year in that transition. We think we're well-positioned. Normally, when we go through an interface change, we can typically get some incremental ASP upgrade. I wouldn't say it's massive, but certainly every time you go through those transitions, even within interfaces from different speeds, we typically add more value, and we can price that into the whole solution.
I think there's some incremental there, and I think we're prepared for it, and we're prepared to support the entire range of latest interfaces and SSD to really go after the whole market.
Thank you. Our next question comes from the line of Christopher Rolland with Susquehanna International Group. Your line is now open.
Hey, guys. I'll echo my congrats on the solid results. One of your chief competitors here has really strong ASIC business. Perhaps you can talk about some of your capabilities that you're going to get with Cavium, some technologies that you're going to be able to add into that business. Can that become a much more meaningful business for Marvell in a few years, call it?
Sure. Yeah, I think we're still early days in thinking about that. Clearly, there's a competitor out there that's got very strong capability in ASIC offering from an IP point of view and just having an established business model and a successful one. That's probably a question that needs to be addressed and answered down the road, and that's a longer-term strategic question about whether when we combine with Cavium and we get all the IP rationalized and in place, is there an opportunity to do more broad, custom ASICs. Today, we do actually engage in ASIC opportunities on our own, but we don't run it as a standalone business. That decision, I would defer till sometime later and after we get the two teams combined. Clearly, it's a capability that we will have. Whether it makes sense or not, is to be determined.
Thank you. Our next question comes from the line of Tom Sepenzis with Northland. Your line is now open.
Hi, thank you for taking my question. I was wondering if you'd give us a little color on what you're seeing in the carrier networking business. It's been a little bit, I think, weaker than enterprise over the last year. Just what are you expecting, or what should we be expecting here over the next couple of quarters from that?
Sure. No, that's, I think very well-known, the issues out there with the operators, with the base station providers, with the infrastructure companies supporting the carrier network. We're not as exposed to that business, which is why I think you're seeing our networking business perform really well through this cycle because of where we're exposed. We do have carrier business. It's very lumpy. We've had a few of our customers this year, as an example, really not do what they thought they could do, and it's not that they tried to overstate their numbers. There really has been weakness. We do see that continuing. I think carrier, in general, has just always been a very lumpy business and very hard to predict.
Typically, those same customers for carrier are the same customers for enterprise, so we try to leverage the design wins to the extent we can across both of those segments. For us, it's not as meaningful a number, so it doesn't really move the needle for us on a quarterly basis. I would echo what you're signaling, which is It's been weak. It's probably going to be weak through 2018, and it's probably going to be a little bit lumpy.
Thank you. Our final question comes from the line of Atif Malik with Citi. Your line is now open.
Hi, thank you for taking my question. Matt, I want to go back to the comment that you made about a NAND supply starting to loosen up earlier than you expected. Assuming that supply continues to loosen into next year, how will it impact your storage business? Particularly on the pricing side, I assume you guys benefited from pricing on the controller side as the market was tight this year. Can you just talk about how the dynamics reverse, if at all, next year with the NAND supply loosening?
Sure. No, understood. I think clearly from a NAND cost per gigabit point of view, more supply obviously is going to drive the pricing down for the NAND. From a controller point of view, even though the supply has been tight, it's a competitive business, right? Each of those sockets we compete for, they're contested, and we've got to sharpen our pencils and be competitive. I wouldn't say that the environment has really, even though NAND supply has been tight, I wouldn't say that we've sort of gotten a pass as a supplier. I think we've had to be aggressive and be on top of it to win business. I anticipate that's going to continue into the next year.
I think to the extent that our customers can really ramp higher volumes, then they can come back to us and say, "Hey, look, you quoted us 1 million pieces and you gave us some price, and now we can actually ship 5 million pieces and we want a discount," then we would probably go ahead and do that. We would also benefit from higher revenue than we're projecting and probably be able to get the supply chain costs to come down commensurate. I guess the long story short, I don't see with the kind of business we have today that there's going to be downward ASP pressure that's going to be more than normal given that supply is loosening. We'll have to see how the dynamics play out.
We're already preparing for if there is higher volume, we want to make sure we're competitive. I hope that answers your question. Typically, the businesses that we're in and the kind of solutions that we provide, we win them early on. We give some kind of longer-term price quote. We give sort of a matrix of volume versus price, and then that's what customers buy to. We don't have the spot pricing effect that maybe the NAND companies themselves go through.
Thank you. With that, I have no more questions. With that, we will draw the call to a close. Thank you all for listening. You may all disconnect.