Good day, ladies and gentlemen, and welcome to the Q4 2018 Marvell Technology Group Limited 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. [Operator Instructions] As a reminder, this conference call is being recorded. I would now like to introduce one of your host for today's conference, Peter Andrew. Mr. Andrew, you may begin.

Thank you, and good afternoon, everyone. Welcome to Marvell's fourth quarter and 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 quotations. 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 turn the call over to Marvell's President and CEO, Matt Murphy.

Thank you, Peter, and good afternoon to everyone on the call. The completion of fiscal 2018 marks my first full fiscal year as the CEO of Marvell, and our performance over this period demonstrates that Marvell is successfully transforming itself into a much more capable and competitive company, one that delivers much greater value to customers, employees and shareholders. Since I joined Marvell 19 months ago, we have increased our focus on our three core markets, storage, networking and connectivity. We've also added greater discipline into our product development process, retooled and reinvested in our sales force and go-to-market strategies and continue to drive efficiencies in our new business model. Our fiscal 2018 results clearly show that these actions are paying off. Fiscal 2018 year-over-year revenue from our core businesses grew 7%. You may recall from our Investor Day a year ago that our stated goal was to grow faster than our core end markets, which at that time were estimated to grow 6%. I'm proud to report that we exceeded that goal. I'm also pleased that this growth was broad-based, with storage growing 8%, networking growing 1% and connectivity growing 14%. It is worth noting that our networking business, excluding legacy, grew 7%, and I'll talk about that in a minute. For fiscal 2018, non-GAAP gross margins increased over five percentage points from a year ago to 61.4%, exceeding our target of 60%. Non-GAAP operating margin increased to 25.9%, up over 11 percentage points from a year ago, and we're making steady progress towards our long-term target of 30%. Finally, in the 10 months leading up to the Cavium announcement, Marvell returned $647 million to shareholders, which is 130% of our full year FY 2018 free cash flow. By all measures, our results were impressive. Now turning to our…

Thanks, Matt, and good afternoon, everyone. I'll discuss the highlights for our fourth quarter and the fiscal year 2018 and provide our current outlook for the first quarter of fiscal 2019. Revenue in the fourth quarter was $615 million, above midpoint of our guidance we provided in November. Our core business of storage, networking and connectivity grew 8% year-over-year. Storage accounted for 53% of revenue and grew 4% year-over-year, driven by the success of our SSD products and continued growth in enterprise and the data center segments with both our SSD and HDD solutions. Networking accounted for 25% of revenue and grew 5% year-over-year, driven by our new switch, PHY and embedded processor product. Connectivity had a solid quarter and accounted for 14% of revenue and grew 31% year-over-year, driven primarily by growth in high-end voice-enabled and home streaming devices. Finally, other product accounted for 8% of revenue, which was about $2 million better than we expected due to last time buy activities. GAAP gross margin for the fourth quarter was 60.7%, and the non-GAAP gross margin was 62.3%, a record level for Marvell and an increase of 4.5 percentage points from last year. GAAP operating expenses were $319 million, higher than our guidance, primarily due to the segment of 2015 security class action claims for $72.5 million during the quarter. Non-GAAP operating expenses was $217.6 million, consistent with our expectations. GAAP operating margin was 9%, and the non-GAAP operating margin was 27% versus 20% a year ago as we were successful in driving [indiscernible] our model. GAAP earnings per diluted share were $0.10, and the non-GAAP earnings per diluted share was $0.32. For fiscal 2018, we delivered non-GAAP EPS of $1.19, up 80% versus the fiscal 2017. Let's now turn to our balance sheet. We ended the quarter with…

[Operator Instructions] And our first question comes from John Pitzer with Credit Suisse. Your line is now open.

Yeah. Good afternoon, guys. Congratulations on a solid result. Matt, maybe the first question from me, just going back to the legacy networking business. Can you help us understand what percent of networking that now represents? And given that you saw push-outs from fiscal fourth quarter into the fiscal first half, can you just give us a little bit more color what drove those push-outs and why you have confidence that they're not going to fall in the first step of the new fiscal year?

Sure. Yes. Thanks, John, happy to answer the question. So the answer to the first part, which is the percentage of networking today that's in legacy is around 15%. It was in - last year, it was in the 20% range. As new products have ramped up, that percentage, that legacy continues to go down, but it's about 15%. And as I mentioned in my remarks, the core portion where we have our new product ramps occurring came in pretty much in line with what we thought. So we saw very strong year-over-year growth there. That was on track. Within legacy and the push-out, it was kind of interesting, it was really driven by really one product area, which was our NPU product area. It was pretty specific and it moved out from Q4 and it'll be consumed within the first two quarters of fiscal '19. And my understanding of the push-out is it's related not so much to in-demand weakness as much as some issues around component supply within that particular customer and that particular box. So it happened during the quarter. We see it selling through or resuming production, but sort of absent that, I'd say, overall, that was pretty close to what we thought. I will say, as a lot of companies have reported, communications has been a bit choppy for everybody. We would sort of get some color that enterprise and wireline has held up pretty well, things that are exposed to carrier and telco and others have been more choppy. But in aggregate, just so everybody understand, I'd say it's primarily driven by kind of one product, one issue, and we certainly see demand recovering on that one.

That's helpful. And then, maybe as my follow-up, on the SSD controller business, I think, one of the questions that I get asked often by investors is the HDD controller market ended up being mostly a merchant-supplied market. When you look at the SSD controller market, guys that are providing NAND or clearly trying to move up the value chain. I think, last week, Western Dig launched a family of SSD products with their own internal controller. Our work would suggest that hyperscale customers don't want to move in that vertically integrated direction, but I'd be curious on your thoughts on how you think the merchant versus the in-house market plays out in the SSD controller market over time?

Okay. No, got it. And I'll give you my perspective. So I'd say, even going back 1.5 years plus ago when I joined, I think, this same issue was top of mind as we started talking to investors about SSD and our plans there. And I think, there's been an effort across a number of the producers of NAND to have their own internal solutions. Some of those have been successful, some of those haven't. We've been operating in this world for some time now, so I think, you should expect, at least for the time being, that you'll see a mix of Marvell solutions in the market, as well as some of our customers doing some other programs internally. And I'd say that's not necessarily a change. Its how we've been operating and we've been able to, I think, execute and grow really well through this cycle. And we certainly still feel very good about our outlook over the coming year. I think, you would sort of would take the words out of my mouth, I think, you're right on where the puck is going, in particular trends we see within the data center is that one of the models is for more control over that design by the cloud hyperscale companies. That's a great partnership opportunity for Marvell to step in and get much closer to the development of the end equipment and then work with NAND partners to enable the whole solution. So that business model is very much alive and well along with our merchant efforts. And I'd end by saying, my job and the team's job is really to continue to innovate in this area, invest ahead of the curve and really eliminate reasons why our customers might even want to think about doing these things internally. And I think, our team is doing a good job at that. So I think that's - it'll be my comments around SSD.

And to add to what Matt just said, yes, the design cycle of SSD product is quite long, right? Really long cycle. So we're actually very encouraged and pleased with all the design wins we're working on. So we don't see any issues there. We are - our team is really focused on the design wins and working on those wins to ramp up in the next several years.

And then, just maybe as a housekeeping. On the accounting change from sell-through to sell-in, what percent of your revenue does that cover? And you made the comment you don't expect a material impact to revenue, was that a fiscal year comment? Or is that also true for the fiscal first quarter? Because we've seen other companies in the quarter where they make that change have a positive revenue impact. I'm just trying to make sure your comment was not just a fiscal year comment, but also for the fiscal first quarter?

Yes. So first, you're right. Marvell's distribution revenue is about 20% of our total revenue. More importantly very significant portion of that revenue actually is demand fulfillment. So when you think about Marvell's revenue really generated from distributors is a very small, probably below 10% of revenue. And typically, our inventory in the distribution channel is about demand fulfillment and to keep the operational flexibility. That's why we don't see much change on our revenue, either for Q1 or going forward.

Perfect. Thank you, guys. Appreciate it.

Thanks. And our next question comes from Christopher Rolland with Susquehanna. Your line is now open.

Hey, guys congrats on a great quarter here. On networking, so a key customer of yours has been talking up their new campus switch product. I think they described it as being the fastest-growing product line in their company history. Perhaps you can talk about what that means for Marvell and that opportunity there? And as you guys size up networking, what do you think the biggest opportunities for fiscal '19 are? Thanks.

Sure. Thanks, Chris. Yes, a couple of comments. So I think that particular announcement was really good for that company. It's going to be good for Marvell. And I think, even maybe on a broader scale, it really, I think, lit a fire within the industry to drive this upgrade cycle in the campus area, which as you've heard over our commentary for the last year, 1.5 years, that's really been an area where we've been very focused on a global basis to drive all of our new products that were refreshed into that segment. So we do think that the enterprise and the campus has a nice tailwinds to it from that point of view, and we do think that, that's going to be certainly positive for us. And so I think, that's going to be one factor as we look into '19 per your question. I think the other one really, I would say, in terms of design win activity, that appears to have really picked up, and it sort of comes on the heels of a lot of the announcements that people were making at MWC. And with, I think, the sort of view that 5G is really alive and well and probably happening earlier than people thought, there's a tremendous amount of design activity going on there. And in our networking product line, in particular, we're seeing very strong demand in our higher capacity switches, including our 10 gig switches, as well as our 25 gig switches that we had initially targeted for the enterprise data center. Those are now seeing very good design win momentum across that market as well. And so when we look out to '19, we see a mix of some resurgence in enterprise and campus, along with some new opportunities and new boxes being built to support the next-generation wireless technologies that are coming.

Great. And on the SSD side of things perhaps. You talked a little bit in your prepared remarks about some hyperscale engagements. Just any progress or detail there? And then, for SSD overall, I think, you said that, that would be exiting this year about 30% of storage. Did you track to that 30%? And how do you see this as a percentage exiting fiscal 2019?

Okay. So on the first one, I would not get very specific other than to say we've had a lot of positive engagement in this hyperscale area for some time. That's continued. I would probably also say that in the meetings I've had now with the Cavium folks, with those types of customers, the level of engagement we're having is even at higher levels, probably within those companies that we've been engaged. And I think it just presents an additional opportunity and validates the fact that, that's going to be a key part of our comprehensive solution sell when we look at engaging deeply in the cloud. With respect to how did we do on SSD as a percent of storage, we set this bogey out there of 30%. We did have a record quarter in SSD, so we are very pleased. We came very close to our 30%. I wouldn't say it was 30%, but it was really darn close. And quite frankly, as we noted in our comments, because HDD also had a very strong quarter for us and performed above our expectations, probably the fact that HDD did a little bit better caused us to come in just slightly below our 30% target. So that's one I really can't complain about because both businesses did better. And we're not setting an official target at this point for ending fiscal '19 at this time, but we absolutely expect that this will continue in terms of the mix shifting from HDD to SSD in an environment where even HDD we can hold flat. So any growth we see in SSD will, as a percentage, will be because of top line, not because we anticipate HDD decline to make up the difference. Hope that's helpful.

Perfect. Thanks, Matt and congrats again to everybody.

Okay, thanks.

Thank you. And our next question comes from Craig Ellis with B. Riley FBR. Your line is now open.

Yeah. Thanks for taking the question and Matt, I just wanted to turn it a longer-term matter. So first, congratulations on the very strong fiscal '18. Great job for the year. Looking at some of the growth rate benchmarks that you talked about, storage of 8%, networking, I think you said 7%. As you look at 2018, what are the gives and takes to growth in the business that would be in that mid to high single-digit range for a second consecutive year?

Sure. Yeah, Craig. So I think, again, we were happy with the performance, as you know. I think that - and we do see the momentum continuing. While we're not giving fiscal year '19 guidance, we do believe that our thesis on storage remains intact and very sound, which is really continue to remix our HDD business and add incremental enterprise and nearline type of revenue as client has a secular decline in it due the remix in SSDs. So net-net, we're still sticking to our strategy, along with preamps, by the way, of keeping HDD flat as a goal. And then, we're coming off a great Q4 in terms of SSD being a record quarter. And we certainly hope and anticipate that, that kind of growth will - that we'll see growth in that business again in fiscal '19. So when you add those two, clearly, you get something that's above flat when you look at growth rates. And we're optimistic we can keep growing that business. In networking, especially as legacy has come down and is a much more manageable portion of the total, I think that gives us more comfort that the strong momentum we see in our new products will have some runway, and the headwind from legacy will be a little bit more diminished than we've had before. And we think we can manage both of those. And so, I think, in summary, what I'd say is we see these two businesses for Marvell standalone, as well as even when we combine the two companies, is really being a strong growth engine in terms of both top line contribution, as well as gross margin accretion. And again, it's on the heels of new products and really important secular trends that are driving the need for more storage, higher bandwidth, data traffic management, more processing of data and more security of data. And I think all these things, we see that thesis playing out and benefiting those two businesses that you mentioned.

That's very helpful. Then the follow-up would be, you made two comments, I thought, were interesting with respect to Cavium. One, the feedback that you're getting from customers. And then, your high confidence in hitting prior - previously stated synergies targets. I'm just wondering if you could elaborate on those a little bit further to help us understand what you're seeing as you get closer to closing that deal? Thank you.

Sure. So I'll comment on the customers, I'll give a point-of-view on synergies and I also have Jean chime in on synergies as well since we've been joined at the hip on this one to make sure that we execute to everything that we've committed. So on the customer side, I've had the good fortune to spend a lot of time with Syed and the team since, really we got this thing announced. And I'd say, what's really resonating are a couple of points, and this would be, I'd say, across a wide range of end customers, including top cloud hyperscale companies, top companies leading in the wireless infrastructure race and also some of these companies that are leading the resurgence in enterprise and campus demand, both wireless and wired. And that is - we're able to come in and basically say, look, we are a scaled up innovation-oriented, fully focused on infrastructure supplier to you. So our primary focus in life as a company is to focus on being an infrastructure solutions provider. And so this scaled up R&D budget that we now have, much more bulked up market cap and revenue provides a lot of assurance to those companies that we have the scale and the strength to invest. We also have the technology, the technical capabilities, the people, the know-how and trusted relationships coming in from both sides. And so it's resonating really well because we don't look like a company that's going through distraction right now. Even though we're going through an integration, it looks like, to our customers, because this has been so positively received by both sides, it's really been a joint effort. So I think, when the customers look out, they see a really new, unique supplier to meet their needs. And I'd…

Yes, I think, it's really two things. Matt mentioned the one, very detailed planning. Second is really accountability of the team to take responsibility. Of course, when you look at the synergy, we're very comfortable with the range of the synergy within 18 months. But bear in mind, a lot of the synergy will rely on system integration and the ERPs. So from the shape of the synergy, you're not going to see much in the first six months. But then, once you get to the integration of the system and the ERP, a lot of the synergy will happen between nine to 12 months. So I would say, we have a very detailed planning and the timeline over the synergy, that's really the comfort level we have about the synergy realization in the next 18 months.

Okay. That's helpful. Thanks, Mart. Thanks, Jean.

Thank you. And our next question comes from Joe Moore with Morgan Stanley. Your line is now open.

Great. Thank you. I'm wondering if you…

Hey, Joe, are you still there?

Yeah. Can you hear me?

Yeah, now we can. You dropped off for a second.

Sorry about that. You talked about the SSD business in the context of demand-supply. Obviously, very tight last year. It seems like it's more balanced now, but still pretty healthy. How does that change the mix of hard disk drive versus SSD? And does that have any effect on your market share and your opportunities in that market?

Sure. Yes. Thanks, Joe, and I'll make some high-level commentary. I think, clearly, happy to share some thoughts, but don't want to be the bellwether or the canary in the coal mine on calling NAND supply and demand. I think, even some comments last quarter may have gotten construed that way. It wasn't intentional, just trying to be helpful. So with that in mind, what I'd say is probably continued trends and - in terms of what's widely publicized, which is that supply coming online, I think. What I said last quarter sort of ended up coming through and then probably will continue. And we think that's a good opportunity for us, certainly, as some of the SSD customers we had last year who were supply constrained can now produce and probably get some or all of what they need. As far as the outlook and the impact on HDD, it's something we monitor. That business so far has performed quite well for us, not only because our end customers have done well but our sort of increased presence in some of these newer areas for us like enterprise and nearline have given us a bit of a tailwind there. So I guess, I'm not in a position or ready to call a bigger picture what's going to happening to HDD, SDD for the year, other than, we think, we see strong growth drivers for us because of our own unique design wins and positioning in both. And we'll, for the time being, stick to our thesis around keep dollars flat in HDD, albeit certain quarters will be a little higher, a little lower. And then, keep really executing well on the SSD growth that's out there, and try to make it happen. So that's the landscape of what I can share at this time.

Okay. That's helpful. And then, I wonder also if you could just touch on your gross margins keep going up. I think it's been a couple of quarters now you said sort of mix is going against you because margins continue to rise. How should we think about that? Is it sort of - is that just simply mix within each of the segments? And is there anything kind of formulaic we can look at as to how much gross margin call through as you continue to grow revenues?

Sure. I actually do think - and it's not intentional, obviously, to sound like a broken record on this one, which is, hey, gross margins did better than we thought, gross margin outlook is better than we thought, but we also have this sort of mix issue going against us, and it has been true. Really, what's happened is connectivity was just a really strong performer for us last year, and it's sort of been that way every quarter. It's done a little bit better than we thought. So that being said, we do think, because of in the sort of what's been realized so far, absolutely has been improved mix within each of the segments. Meaning you've gotten higher value, design wins or new product ramps in storage and networking or in connectivity. Despite the fact that, that business did grow year-over-year, we have been actively remixing quite a bit in that business in terms of letting applications just sort of go away and really focusing on areas where we could make a gross margin impact and really focusing on profitability versus top line. But top line also worked well there. So what I'd say is I'd stick to my same story, which is we do think we certainly - we gave, I think, a solid guide for Q1, which was 62 to 63 on the GM. I don't think anybody a year ago would have been sitting here thinking, when we are at Analyst Day 2017, that we'd be talking about guiding 62 to 63 a year later, so we're really happy with that because of the way we've managed the company. But beyond that, while we don't give specific ranges, our seasonally strong quarters and storage begin in Q2 and Q3. And overall, kind of bigger segment mix should help us there. So we'll see. We take it one quarter at a time, but directionally, certainly, we're very pleased. And I would also just give a quick shout out as well to our manufacturing and operations team who continue to do an outstanding job really working with our supply chain partners and really optimizing our cost structure, so we can both grow revenue, reduce cost and make us more competitive with - in the market.

And then, of course, when we combine with Cavium, it will further expand our gross margin. And as you recall, when we announced the deal, we did say the combined company target gross margin will be 65%.

Great. Thank you very much

Thank you. And our next question comes from Blayne Curtis with Barclays. Your line is now open.

Thanks for taking my question. I'd like to congrats on a good quarter. I just wanted to follow up on connectivity. Maybe if you could just talk about why the fall in seasonality? Is it inventory issue or just an earlier build? And then, you've telegraphed that mix away from some non-strategic business. Just kind of as you look at this year, is there any kind of lump in that? Or is that kind of a gradual thing as you get new product and you let some walk out the door. I just want to understand, as this year develops, new business coming in versus you're walking away from?

Hi, Blayne, this is Jean. I'll take the first question about the Q1 guidance. Yes, when we guided mid-teens sequential increase, it's better than normal seasonality, largely because of the one particular customer, they are actually building product in Q1 versus typical building Q2, Q3 for holiday season. So it's very unusual situation. So I do think Q1 revenue for wireless is unusually high. And Q2, Q3, you're actually going to see some kind of a decline more than typical for this particular customer, and they - which is a large customer. So that's the dynamics of Q1 versus Q2 and Q3. As far as the strategy wise, overall, we are really trying to focus on profitability of the Wi-Fi business. So we're certainly trying to get out some of the low-margin applications in the consumer side of the business. Matt, you probably can add more on the strategy side.

Sure. I think, we've - just to reiterate it for all the investors on the call, we've - as I said, we've focused on, really, profitability in that segment to make sure it's a contributor on certainly gross and an operating margin. And we've made huge strides in our wireless business. I think, this question you got, Blayne, about what does that look like as consumer, maybe lower margin or lower value add business rolls off. I mean, that's why, even in the last year, we've been very judicious every quarter about how we think about that business and guide it because there has been a big remix that's been ongoing and will continue to be ongoing probably over the next year. And as Jean mentioned, because of one of our customers, I think, doing a ramp that has some of what would probably be in Q2 and a little bit of Q3 in Q1 that probably linearizes more over, let's say, 2.5 quarters versus two quarters. But overall, I mean, there's been enough sort of new stuff coming in and old stuff rolling off. And we're not - beyond that, I think, that's as much as I can be helpful other than we're not - we just can't guide Q2, Q3 and beyond in detail just yet.

And then, just on the gross margin, you - I would think that the wireless would be a headwind to gross margin, you're still growing it, so maybe just talk about the drivers to gross margin.

Sure. So again, I think, we have the headwind issue. We actually guided to seasonality, obviously. Slightly lower revenue. Of course, it was, I think, still a solid guide. But again, very pleased to see on the GM, and I think, as I said, I think, we see new products ramping up. I mean - so if you think about networking, which has been a story we saw in Q4, and we certainly said we're going to see it in Q1, those new products have been much higher value and much better margin than historical Marvell legacy - or Marvell networking. Also legacy is lower than the new stuff, so as that mix shifts within networking, that's helpful. And again, I'd say, its continued execution on the cost reduction side as well, which has been a benefit to the entire company, to the entire enterprise. And I'd just reiterate back to what Jean said when we think about integrating Cavium, all these fundamental improvements with how we manage our supply chain and operations and manufacturing overhead, all that benefit we see is just translating right over as we think about synergies in integrating Cavium. So to the extent we're getting improvements in our own margin, not only does that help what will be the blended margin of the combined company, but those structural improvements, we think - we very much think, we can pass on to the Cavium side. So I'd just say it's a mix of things, Blayne. There's no one trick pony on it in terms of why it's going up, other than we've been really focused on it. And ultimately, the only way you really do this over time is we believe strongly, gross margin is the key indicator of the innovation and the value you're delivering. And the fact that our new products have been ramping strongly across our different segments, it's helping our GMs. And that's sort of the - that's the story.

All right. Thank you.

Thank you. [Operator Instructions] Next is Vivek Arya with Bank of America Merrill Lynch. Your line is open.

Thanks for taking my question. Matt, one more on the storage outlook for Q1, so down year-on-year. Conceptually, SSD growth should be more sustainably offsetting HDD declines by now. Is there something specific in Q1 that is preventing that? And should we assume that storage starts to re-grow year-on-year from Q2 onwards?

Vivek, this is Jean. So when you look at Q1 from a seasonal perspective, HDD typically decline actually 7% to 10% sequentially. So that's really HDDs decline much more in Q1. And in SSD, we're actually going to see year-over-year growth in Q1. It's just that we have a very large base on HDD business. That's why we're not offsetting Q1 particularly this decline. As I said, quarter-over-quarter may be different. We do expect going forward, especially for the fiscal year, SSD growth is going to be faster than SSD market, which is the market is - everybody's consensus now is over 20%. So from that angle, you can see what Matt mentioned earlier is, for the fiscal year, we'll largely keep the HDD revenue to be more consistent to flattish for the year, and SSD growing much faster than market. So I think, quarter-over-quarter, there may be some volatility. But overall, we feel pretty good about the prospect of the future.

And anything on SSD controller competition? I know you mentioned that if you set aside competition or in-sourcing from the customer side, has the competitive landscape with other merchant suppliers changed in any way? Or do you think you still have pretty strong competitive moat in that area?

Sure. No, I think, we've certainly had competition in this area for some time. It's - we take nothing for granted. And the competitors traditionally we've had over the last couple of years, we still have them. We're still on those, battling out on these opportunities. So I think, the competitive landscape, I wouldn't say, has changed, but I would say that as our scale has increased in terms of just the breadth of customers we're supplying, the types of end applications that we're in, which range all the way from consumer end client all the way up to the most advanced hyperscale and enterprise applications in the world, that scale has allowed us to continue to invest in that business, I mean, in terms of adding R&D resources and expanding our abilities to develop more products and more IT. So we're becoming, I'd say, increasingly more relevant than we were 1.5 years ago just because that business scale is now at a point where we can really - we have a very comprehensive set of solutions. And as I pointed out in my prepared remarks, despite the fact that we've had all this great run, we're still ramping new customers. This past quarter, we ramped a brand-new Tier 1 NAND company that we hadn't had before. We ramped also in a new cloud hyperscale application that we weren't in before. So that's all new and incremental on top of what we've done. So that's what I'd say. I think, it's never going to be a pushover between internal competition, merchant, ASIC, which we compete in as well as some other folks. But again, from having a breadth of portfolio, scale of revenue and R&D investment, we do believe we have a significant competitive moat there, and we're head and shoulders above the others in the market just because of that scale.

Thank you.

Thank you. [Operator Instructions] Our next question comes from Ross Seymore with Deutsche Bank. Your line is now open.

I wanted to go back to the networking side. Matt, I know you said what the issue was in the fiscal fourth quarter about the push0out. But I wanted to think, as you think about it for this year as a whole, I would have thought you'd be a little better than flat in the near term guide, or flat to slightly up if that push-out from last quarter started to help this quarter, given the strength in the end market and some of your big customers. So in the near term, can you just talk about what the puts and takes are for the April quarter? And as we think through the year, where do you think that legacy business will exit this fiscal '19 versus where we just exited the prior year? Thanks.

Sure. Maybe I'll take the second part first, and Jean, I don't know if you want to comment, or we can both do this one together. But on the second part, we don't have an exact number, Ross. But I'd say, it sort of went from 20% down to 15%, probably normalizes in that area, right? The 15 percentage that's left is really pretty long life cycle stuff. So while it will decline over time, it's not - I think, the slope on the downside has kind of moderated. So 15% or a little bit lower is kind of the range. In the near term, Jean, I don't know if you want to comment or?

I think, if you're talking about Q1, right, our guidance actually is mid-single digits or high-single digit year-over-year growth. So it's really primarily driven or led by the switch processor and the PHY portfolio, the refresh portfolio. The legacy is more likely - we're actually cautious because the component has shortage challenge we had last quarter. We're actually very cautious about the legacy side and kept it really at kind of sequentially decline kind of way. We'll see how the legacy perform. But as Matt said, because the shortage is really not because end market demand, but it's just because of some memory component shortage. We do think later on that the year will recover, largely keep the legacy business flattish. It has a long tail.

And I would add, too, if you look at the way we phrased it, and clearly, Ross, I was certainly disappointed just because, I think, overall, that networking business really did well for us in Q4, absent this issue. And that's why we're being very sort of specific about calling it out, but it is what it is. But if you noticed, what I really said was we expected a recovery in the first half, and that's exactly, to backup Jean's point, just being pretty cautious on that one and just not assuming that there's really about - normally, every push-out is going to have - is it one quarter, or pushed to the other quarter, and that's when it comes back. So I think, we've sort of been very thoughtful about that piece of it. But absent that, I think, the rest of it is still working pretty good for us.

Yes. And the other one, certainly, is our PHY business is actually quite significant. And most of you guys probably know, PHY business is actually declining high single digit typically. So we actually will be able to offset that seasonal decline for PHY business in Q1. That's another factor. It's a little bit long getting to the details now, but that's the dynamics of our networking. We actually have a portfolio of business.

Thank you. And our next question comes from Harlan Sur with JPMorgan. Your line is now open.

Good afternoon and congratulations on the solid results. The Marvell team has had great success with Google. I think, your ARMADA multi-core processor and your Wi-Fi/Bluetooth combo chip powers every Google Home, every Google Home Mini and home nox [ph] product. And as we all know, these products have been doing pretty well in the connected home market, and it's a pretty important initiative for Google. So the question is, does this help driving the strong connectivity growth in the April quarter? And then, how can - or how is the team going to leverage its dominance in the connected home part of this customer to potentially getting more silicon content within the data center segment of their business?

Sure. Well, Harlan, I give you credit for asking a very specific question. But as you can probably imagine, and as you've seen sort from our MO, we prefer to not really get into specific customers and specific content and kind of talk at that level of granular detail. From my days as the Sales VP when I was at Maxim, I just learned that customers really don't appreciate it when you start talking about their business, in particular, with seasonality things and other things that you're asking. So can't really comment on it other than to say that, that particular company is obviously of strategic importance to us, both in that particular product lineup as well as, I would say, even of equal or more strategic relevance is obviously becoming a bigger supplier to them in the data center side of their business. But I'll leave it at that. Happy to answer another question, but can't talk about individual customers.

Great. Well, then, maybe I'll just follow it up with just a very quick question. Great job on driving the OpEx lower in fiscal year '18, Jean. How should we think about the OpEx trend beyond the April quarter here in fiscal '19?

Yes, that's a good question. I would say, the way to think about OpEx is if you look at it toward the end of fiscal '18, our OpEx run rate is around the $208 million per quarter. So when you think about 2019, it's really you have to factor into about 3% to 4% inflation, including merit increase, right? So $250 million for Q1, that includes the payroll tax, but then Q2 will have merit increases. So I will say it's around this level for fiscal '19.

Perfect. Thank you.

Thank you. And our next question comes from Karl Ackerman with Cowen. Your line is now open.

Hi. Clearly, hyperscale CapEx is driving not only higher demand for networking, but also storage. I'm curious, do you embrace the view that nearline hard disk drives remain capacity constrained in the first half of 2018? And if so, how is that impacting your ability to both service the hard drive or the HDD vendors as well as receive premium ASPs in this environment?

Okay. So appreciate the question. I think, though, the best guy to ask would be the company that we're supplying to. I think that, that one, they clearly have much more insight than I do, and there's only three of them to ask, so I would probably err on that side. So - and then, with respect to specifics around ASPs and things like that, I mean, all we have said historically is, at a high level, as we penetrated into enterprise and nearline in HDD, we certainly have a higher ASP per drive that's a client. But that's about it. So I don't really - not in a position to get into specifics about that one or kind of what my end customer dynamics are going to be in that sort of very specific situation you mentioned, Karl.

Thank you. And our next question comes from Mark Delaney with Goldman Sachs. Your line is now open.

Good afternoon. Thanks for taking the question. I was hoping to follow up on the comments about distribution. Jean, I think, you said it's about 20% of Marvell's revenue currently. I believe that at the end of last year, Marvell announced a global distribution agreement with Avnet. So maybe you can help us better understand what you're hoping to achieve in distribution and the reason for the new agreement? And once - or if Cavium closes, how are you thinking about using distribution going forward as a combined entity?

So I'll clarify the percentage of revenue and Matt can answer that other question. So we have about 20% of our revenue into distribution revenue, but Marvell's distribution revenue, a very large percentage is demand fulfillment. And of course, we're trying to get more demand creation from distributors, but I just want to make sure we understand the numbers first.

Yes. And then, yes, you're right, Mark, to note that we did make a change in our distribution network, not only with Avnet, but we've made a number of other changes. And really, I would put this one in the bucket of sales force and go-to-market transformation that, really, Tom Legatta has been driving since he joined. Soon after he joined, he and I both sat down and just took a look at the global landscape. And given that it's 20% of our business, we quite frankly just had way too many distributors scattered all over the globe. There was not a comprehensive global distribution program, design registration management was all over the map. The way we were incentivizing those distributors from a margin point of view was not consistent. And I would say we basically had a whole bunch of distributors without a program. And that's why, as Jean points out, it's only 20% of our revenue, and of that, maybe half of it's actually been created for distribution. So that to me seemed way too low. Tom's reaction was the same. We subsequently added an executive to run distribution in last fall. And then, toward the end the year, we announced a series of changes, one of which was pointing Avnet as our global distributor and really expanding that franchise worldwide. We terminated a number of others and really got a much more focused channel strategy in place that, I think, for both sides, both distribution and Marvell, actually optimized the incentive structure, and certainly made it much more simple for us from a go-to-market standpoint. And you should expect the same playbook that we've implemented on Marvell, we're going to take a very close look and already have with Cavium's distribution network and the activities that they have going on there. By the way, they have some very nice activities going on within distribution and with Cavium, so I think there are some, probably, some top line synergy there to take advantage of that. So we're going to put those two together and you should expect we'll have a very professionally managed, efficient global distribution plan implemented as part of our sales track when we think about integrating Cavium.

Thank you. And we have time for one more question. Our next question comes from Harsh Kumar with Piper Jaffray. Your line is now open.

Yeah. Hey, guys. Thanks for squeezing me in. And congratulations. So first question for you, Matt, is would you be able to give us some sense of how much the preamp revenues are? When they become significant? And given that you're hopefully going to get that revenue to be significant, hopefully, this year, why - is it fair to assume that the HDD business could potentially grow in absolute dollars versus your commentary of flattish?

Sure. So just to give you a sense of where that business is at, while we're not dollarizing it at this time, think of it as two years ago, we had no revenue, products and design. Last year, we started shipping in some kind of low volume preproduction drives that were out there. We see this year as a year where we're going to probably see some of those drives go into full volume production, but that would be in the back half. And you should assume that, that level of revenue is not significant. It's going to take time because the way that these preamps work is you really design them fairly specifically for an individual customer and an individual drive program. And those - our strategy has been obviously is try to get those pair up in applications where we have our own SoC. So that's going to be just kind of an ongoing journey that we'll be on. And we're not breaking out the amounts. And certainly, if you talk to my team that runs that entire HDD business, they're obviously - their strategy and plan is to try to - when I say flat, that's how we're modeling it as a company. Obviously, their plan is to try to grow that business by getting more share, getting more ASP, getting preamps to ramp. And so it's not like we're trying to limit ourselves there, but we will kind of stick by our modeling, our models that we indicated earlier on the call, which is preamp ramping is one of the aspects that allows us to keep HDD "flat" over the year. But we're not really able to get into the numbers. But as we make progress, you should assume, just like everything we've been doing, we'll be more transparent, we'll give you more details. Once it's more in front of us, and I think I've been served well by not trying to project too far out on new ramps, I think, at the end of the day, it's very hard to do that accurately. And I wouldn't want you to over or underestimate too much. So I think HDD flat is probably a good way to think of it with preamps being part of the equation.

Thank you. And I would like to turn the call back over to Peter Andrew for further remarks.

Okay. Well, thank you, everyone, for joining us today. And we look forward to talking again next quarter. Thank you. Good evening.

Ladies and gentlemen, thank you for your participation in today's conference. This concludes the program, and you may now disconnect. Everyone, have a great day.