邁威爾科技 (MRVL) 2020 Q1 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the First Quarter 2020 Marvell Technology Group Ltd.

Earnings Conference Call.

(Operator Instructions) As a reminder, this conference call is being recorded.

I would now like to introduce your host for today's conference, Mr. Ashish Saran, Vice President of Investor Relations.

Sir, you may begin.

Thank you, and good afternoon, everyone.

Welcome to Marvell's First Quarter Fiscal Year 2020 Earnings Call.

Joining me today are Matt Murphy, Marvell's President and CEO; and Jean Hu, our CFO.

I'd like to remind everyone that certain comments today may include forwarding 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 refer 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, I'll turn the call over to Matt for his comments.

Matt?

Great.

Thank you, Ashish, and good afternoon to everyone on the call.

Before we discuss our recent performance and outlook, I'm going to walk you through the 3 significant portfolio changes we've recently announced and put them into strategic context.

As you may recall, when we began our transformation back in calendar 2016, we restructured Marvell, pivoting to a new strategy with an intensified focus on storage and networking as our core markets.

We put in place a number of key initiatives to organically improve margins and ignite top and bottom line growth.

We delivered significant improvements from these initiatives in 2017 with substantial margin expansion and revenue growth.

In November of 2017, we announced our intent to leverage our strong balance sheet and share price to fund the transformational acquisition of Cavium.

We closed this highly complementary merger midway through 2018, adding processing and security to our portfolio, along with additional scale and a more complete networking technology platform.

Since then, we have successfully completed the integration of Cavium, raised our synergy target over our initial estimate and driven multiple new high-growth vectors for the combined company in 5G, automotive, ARM servers and AI.

And as you are well aware, 5G is rapidly emerging as one of the most exciting growth catalysts for Marvell.

While we are on this journey primarily focused on execution and integration, we continue to closely monitor our end markets and the broader semiconductor industry.

We recognize that the market was softening late last year and our view is that after many years of growth, our industry was bound for a contraction, particularly against an increasingly uncertain geopolitical backdrop.

Nevertheless, we determined that despite the challenging environment, we would sustain our R&D investments and continue to collaborate closely with customers in each of our growth markets.

And rather than just wait out the down period, we continue to assess the external market for inorganic opportunities and systematically managed our current portfolio to accelerate our pivot to infrastructure solutions.

This culminated in the 3 transactions we recently announced, namely our planned acquisitions of Aquantia and Avera as well as the divestiture of our Wi-Fi business.

I'd like to comment on each specifically.

First, Aquantia.

This is a highly complementary acquisition, which upon closing, will immediately add scale and breadth to our PHY business, while also accelerating our reach into the fast growing multi-gig enterprise and automotive Ethernet areas.

Marvell has been very successful with copper PHYs at speeds of 1-gigabit and below, while Aquantia has been a leader in higher-speed solutions of 2.5-, 5- and 10-gigabit.

And as we look longer term, with advanced driver assist and autonomous driving continuing to gain momentum, we expect that our processing security and AI capabilities will add on to our automotive networking platform to establish Marvell as a leading supplier to the connected car of tomorrow.

Second is our acquisition of Avera, originally part of IBM's microelectronics business.

This acquisition was motivated by the growing list of requests from our customers for new custom products, including full ASICs.

Avera, with its proven track record in ASIC design, was the ideal partner to complement our standard products and semicustom capabilities.

We have immediate opportunities in 5G, which we can now transition into design wins.

Marvell's broad technology platform will also enable the Avera team to help us become a leading-edge ASIC provider to a wide range of networking and cloud customers.

Third, we're equally pleased about the planned sale of our Wi-Fi business to NXP.

Marvell has a 2-decade history delivering innovative Wi-Fi and Bluetooth products and this business has been transformed over the past 3 years to focus on high-performance solutions, including next generation Wi-Fi 6 products.

Recently, we were approached by several companies interested in acquiring this business to help leverage their scale in the fast-growing connected consumer and IoT markets.

Unlike Marvell, these companies already possess both microcontrollers and a broad distribution channel critical for these markets but were missing Wi-Fi capabilities.

As we evaluated these opportunities, it became clear to us that our high performance conductivity assets would yield much greater economic value at a company with a strong microcontroller franchise than they would as part of Marvell.

This resulted in our agreement to sell this business to NXP.

Fundamentally, the transaction value is significantly higher than what we could have extracted from this business over time on our own.

While these 3 announced transactions are significant independent events, they accelerate our strategic transformation into a leading supplier to the infrastructure market, and they are a reflection of how we actively manage our overall portfolio and allocate capital to invest in the markets where we can generate the highest returns.

Let me just put this into perspective.

Once we complete all 3 transactions in the first year post close, we will be replacing approximately $300 million in Wi-Fi revenue that we are selling for a 6x revenue multiple with greater than $400 million in combined in Aquantia -- combined Aquantia and Avera revenue we are purchasing for a 3x revenue multiple.

This new revenue will be on a faster growth trajectory and will accelerate Marvell's ability to capitalize on 2 critical industry trends, 5G and connected cars, and will also command higher growth and operating margins over the long term.

Proceeds from the sale of the Wi-Fi business will pay for both Aquantia and Avera, and will net us at least $500 million in cash to add to our balance sheet.

I'm also equally excited about the tremendous engineering talent and capabilities these 2 acquisitions will bring to Marvell.

Hiring talented engineers in our industry is highly competitive, so being able to add entire R&D teams with well-established track records of execution enables us to capitalize on market opportunities much faster than building teams 1 hire at a time.

We recognize that these are bold moves, and I'm very pleased that we are able to negotiate and announce all 3 deals in such a short time frame.

Over the next year, we will be focused on integrating these new businesses and successfully transitioning our Wi-Fi business to NXP, while ramping our newest design wins in 5G, data center, automotive and enterprise applications.

Now let's shift gears to discuss our recent performance.

Revenue for the first quarter was $662 million, above the midpoint of our guidance range as our storage revenue was higher than anticipated.

Notwithstanding this result, we recognize that the recent increase in trade tensions, including U.S. government export restrictions to a Chinese OEM, has further dampened demand in an already challenging macroeconomic environment.

These factors are reflected in our outlook for the second quarter.

Specifically, with respect to the customer export restriction, our direct revenue exposure is approximately a mid-single digit percentage of our total revenue primarily in our networking business, although there is an impact on storage revenue as well.

The export restriction was implemented in the second week of our quarter, limiting revenue from that customer to shipments during that short window.

In addition, there are indirect impacts to our business, which we cannot quantify to the extent that some of our other customers' products may incorporate Marvell's chips inside, could also be impacted by the export restriction.

While these issues are obviously beyond our control, we will continue to monitor the situation closely.

Now moving on to the performance of our 2 core businesses.

In the first quarter, our networking revenue was $341 million, down 12% sequentially.

As you may recall, we had expected a double-digit sequential decline reflecting seasonality, tight inventory control at key customers and a cautious outlook from our Chinese customers.

Most product lines decline sequentially, which is not unusual for this time of the year.

In addition, we saw the start of an expected decline in demand from our base station customers as they prepare to transition from 4G products to 5G later this year.

During this transitional period, they are also shipping pre-5G base stations, but we do not see the full benefit from these shipments as OEMs are using FPGAs for processing in the interim.

Our 5G solutions, including our OCTEON embedded processor and Fusion baseband processors, remain on track, and we have now successfully taped out and sampled both of these products.

Our lead customer is set to ramp production on schedule in the fourth fiscal quarter of this year.

As I mentioned in our last earnings call, we secured a second Tier 1 OEM design win for our Fusion baseband processor, and I'm pleased to report that we are on schedule to sample this semicustom product in early calendar 2020.

Now I'd like to share some new developments for Marvell in the 5G market.

As you may know, modern macro base stations are comprised of 2 separate sections, the baseband unit and the radio head.

Today, our processors reside only in the baseband unit, with embedded OCTEONs managing control and data playing functions, while Fusion basebands encode and decode data signals.

However, as we have continued to closely partner with our customers, they have asked us to solve power and bandwidth challenges across the entire base station and not just in the traditional baseband unit.

I am very pleased to announce that 2 of these opportunities have now translated into new design wins.

But first, it's a fronthaul interface solution that connects baseband units to remote radio heads in 5G deployments.

The substantial increase in bandwidth in 5G is driving the need for efficient protocol translations for transport of radio data over fronthaul lengths.

We expect this custom shift to ramp production in the second half of calendar 2021.

Second and even more exciting, we now have our first design win for our processor in the 5G remote radio head itself.

In 5G, there is an increase in the use of massive MIMO to enable higher capacity and throughput.

The complex beamforming needed in these deployments is generating new demand for our optimized low-power processing capabilities, enabling us to capture entirely new content within the radio head.

To meet the fast deployment time line for 5G, customers will go to production with our existing Fusion processor by the end of this fiscal year.

On top of this, customers have also engaged with us to architect custom purpose-built solutions for massive MIMO processing.

This trend greatly expands Marvell's addressable market.

Both of these strategic design wins reflect Marvell's expanding beachhead scenarios that are currently dominated by FPGAs.

Driven by the increase in processing requirements in 5G, including in the radio head, our expertise and power efficiency in processing is opening new, significant opportunities.

Together, we estimate these wins could add, on average, approximately $500 of new Marvell content per base station, assuming a conservative 20% attach rate for massive MIMO in 5G.

Industry analyst forecasts are actually predicting higher attach rates within the time frame we expect to start shipping our solutions.

These are just 2 examples from our expanding pipeline.

While we have been organically increasing our R&D headcount, including the recent opening of our Raleigh design center, the addition of Avera's talented team and extensive custom design expertise will further accelerate our ability to capitalize on these growing opportunities.

Moving now to our second quarter outlook for networking in fiscal 2020.

After taking into account the export restriction, we expect revenue to be down slightly on a sequential basis.

Without the export restriction, we would expect networking revenue to be up sequentially.

Turning to our storage business.

Storage revenue for the first quarter came in above our expectations at $279 million, declining 12% sequentially.

This expected sequential decline was due to a combination of seasonality and excess inventory in the supply chains of our storage controller customers.

Within the quarter, fiber channel revenue was weaker than expected, but this shortfall was more than offset by higher-than-expected revenue from storage controllers.

Our fiber channel business was impacted by a very soft server market in the first quarter, with server units declining by about 18% sequentially, much worse than typical mid-single-digit seasonal declines.

It appears that the weakness in compute demand extends beyond just the cloud and is now showing up in the enterprise as well.

In our storage controller business, while we benefited from better-than-expected revenue, we do not think that this was indicative of any real resurgence in demand from the end market.

We believe that the upside was primarily driven by a customer building buffers to manage a factory transition.

We have taken this dynamic into account in our revenue projection for next quarter.

So for the second quarter of fiscal 2020, we expect that the demand for Marvell storage controllers will remain soft, reflecting continued weak macroeconomic conditions.

After incorporating the impact of the export restriction and accounting for the customer factory transition I just discussed, we expect an approximate mid-single-digit sequential decline on our storage second quarter revenue.

Looking at the first 2 quarters of fiscal 2020 collectively and after adjusting for the export restriction, storage revenue is tracking close to what we would expect, given the current market conditions.

We continue to believe that we will be under-shipping the end market for our storage controllers through a significant part of this fiscal year.

That said, our customers continue to signal a strong demand environment in the second half of the year.

In the meantime, we remain focused on efficiently managing our storage business and continuing to pivot this business towards the enterprise and data center markets.

In closing, despite a challenging macroeconomic environment, we are executing to our strategy and taking bold steps accelerate our transformation.

With the production ramp of our first 5G products on track for later this year and expanding pipeline of design wins, we are well positioned to realize our vision of becoming a leading supplier of infrastructure semiconductor solutions.

And with that, I'll turn the call over to Jean for more detail on our recent results and our outlook.

Thanks, Matt, and good afternoon, everyone.

I'll start with a review of our financial result for the first quarter and then provide our current outlook for the second quarter of fiscal 2020.

Revenue in the first quarter was $662 million versus our guidance of $650 million at the midpoint.

Networking represented 52% of our revenue in the first quarter with storage contributing 42%.

Other products accounted for 6% of revenue.

GAAP gross margin was 54.6% and the non-GAAP gross margin was 64.1%.

GAAP operating expenses were $383 million.

Non-GAAP operating expenses were $295 million, at the low end of the guidance range provided in March.

We continue to tightly manage our [OpEx] in a difficult environment.

I'm also very pleased to report that we successfully integrated Marvell and the Cavium ERP systems early in the first quarter.

With ERP integration complete, we expect to achieve our OpEx synergy goal from the Cavium acquisition in the upcoming second fiscal quarter, 6 months ahead of our plan.

GAAP loss per diluted share was $0.07.

Non-GAAP earnings per diluted share was $0.16 at the top end of our guidance range.

Now turning to our balance sheet.

In the first quarter, we paid down $50 million of our long-term debt and returned $89 million to shareholders through $50 million in share repurchases and $39 million in dividend.

We exit the quarter with $572 million in cash and cash equivalents and a long-term debt of $1.7 billion.

As Matt mentioned, through the 2 acquisitions and the 1 divestiture, we have actively managed our portfolio to maximize the shareholder returns.

We currently plan to use the expected cash proceed of $1.76 billion from the sale of our Wi-Fi business to pay for both the Aquantia and Avera acquisitions.

The resulting excess cash is expected to be at least $500 million.

In addition, we expect our business will continue to generate a strong free cash flow.

As a result, we'll have significant financial flexibility to pay down debt over the next year to achieve our target gross debt to EBITDA ratio of 1.5x.

And also return a large amount of cash to shareholders through share buyback, while maintaining our current dividend level.

As a reminder, we currently have approximately $900 million in outstanding share repurchase authority.

Let me now move on to our current outlook for the second quarter of fiscal 2020.

This guidance takes into account our estimated revenue impact from the U.S. government's current export restriction to one of our customers.

We expect our revenue to be $650 million, plus or minus 3%.

Our expected GAAP gross margin will be approximately between 53% and 54%, and non-GAAP gross margin will be in the range of 63% to 64%.

This projection for gross margin reflects a weaker product mix due to impact of the export restriction, relatively lower storage revenue and a seasonably strong consumer base to product rent.

We do expect our non-GAAP gross margin to return to about 64% in the following quarter.

We project our GAAP operating expenses to be between $370 million and $380 million and the non-GAAP operating expenses to be in the range of $285 million to $290 million.

We expect net interest expense to be $90 million and the non-GAAP tax rate to remain 4.5%.

We anticipate GAAP loss per diluted share in the range of $0.09 to $0.05, and the non-GAAP income per diluted share in the range of $0.13 to $0.17.

Operator, please open the line for questions.

(Operator Instructions) Our first question comes from Vivek Arya with Bank of America.

Congratulations on the Wi-Fi transaction, very impressive.

Matt, does that exit require China approval?

And as part of that, would be very helpful if you could help us quantify the revenue and gross margin and OpEx impact from a fiscal '20 or '21 perspective, so we can focus on your base business?

Sure.

Yes, hi, Vivek.

So on the first question, the transaction on Wi-Fi does require China approval.

And then, Jean, do you want to comment on the second part?

Yes.

So Vivek, on the revenue side for fiscal '19, this business revenue is approximately $300 million.

For fiscal '20, it's about the same level.

The gross margin is around 50%.

I think the operating expense, the operating margin within Marvell is much lower than our corporate average, so it's around 15% to 20% of operating margin.

So that's the way how you can think about this business.

Got it.

And then from a bigger picture perspective, at your Analyst Day, you had outlined very strong operating model of high-single-digit kind of sales growth and over 65% gross margin and high operating margins.

Given all the transactions you have announced, Matt, is that still the operating model?

Or should we think about something else at this stage?

No.

I think you should think about that as the operating model.

There's a number of moving pieces, but part of the comprehension of doing all 3 of these was that we would maintain the operating model that we had committed to.

And certainly, Jean mentioned having Wi-Fi coming out and these 2 others coming in, plus our -- yes, we're planning on doing with it, we're going to stick to the same operating model.

And our next question comes from Ambrish Srivastava with BMO.

I just had a question on -- Jean, this is for you, so that we are all in the same page.

Would you rather that we take out the mid-single-digit Huawei exposure from our model?

And if the answer to that is yes, how should we think about modulating OpEx?

And I guess, gross margin is going to be lower by 100, 150 bps based on the mix, Cavium mix to Huawei.

Then I have a quick follow-up.

Yes, that's a good question.

So on the Huawei side, it's mid-single-digit impact.

So in Q2, we have assumed no Huawei shipment for the remainder of the quarter and going forward.

So from that perspective, Huawei revenue is completely out after the ban of the government.

And when you think about, from the operating expense perspective, right, first, we have to invest in our portfolio and the opportunities to address across all different product lines, different regions.

So Huawei is only a one, like a mid-single-digit revenue customer, so we don't think this Huawei ban will impact our investment level.

And of course, we're very mindful, we are managing discretionary expense to make sure we're disciplined to manage OpEx.

So that's the way how we are thinking about it, that we're going to manage through this Huawei issue and to make sure we drive the long-term revenue growth and the margin expansion.

Okay.

And for my follow-up, real quick, Matt, I just wanted to make sure I understood this.

On the storage side, help me, has the outlook changed versus what you are thinking in Q1, or it's -- sorry, when you talked to us last time?

Or is it about the same in terms of -- if you had said that time that you were shipping below and demand as well, but has anything changed or it's kind of in line with what you're expecting?

Yes, Ambrish, it's very much in line from where we were a quarter ago.

Obviously, we had some outperformance in Q1 that was relative to this factory transition I mentioned.

So no update other than, if you think about kind of Q1 plus Q2, it's in line with what we were discussing the last time we talked.

And we do believe we're still going to continue to under ship demand in Q2 as well.

Yes.

On the gross margin side, maybe what I'll add to the question you asked on the gross margin side is, if you look at our Q2 gross margins, there are few headwinds that impact gross margin.

The Huawei ban certainly has some impact.

But also, we do have a seasonal strong consumer-related product ramp, which actually has a bigger impact.

So when you look at that going back to Q3, that is -- seasonal impact is going to come down from a mix perspective, so we do expect our gross margin to go back about 64%.

And as Matt mentioned earlier, we're not changing our long-term target model.

We continue to believe we'll be able to drive our gross margin to go up going forward.

And our next question comes from Harlan Sur with JPMorgan.

Sorry, can you hear me now?

Yes.

Go ahead, Harlan.

Okay.

Thank you.

Assuming that the Huawei ban continues for the remainder of the calendar year, off of the Q2 guide, I would assume that the team would see sort of normal seasonal growth in Q3.

And then if you do start shipping 5G to your lead customer and some continued inventory improvements on the storage side, you should probably see some sequential growth in Q4 as well.

Is that kind of the right way to think about the overall trajectory of the business for the remainder of the year?

Yes.

I think you're absolutely right.

I think as a simple way to look at it is, Huawei is about mid-single-digit of our revenue.

So when you look at your model, the seasonality for Q3, Q4, the simple way to do it is just take out mid-single-digit revenue out from the ramp, the typical seasonal ramp from the rest of the quarter would be fine.

Okay.

Great.

And then on the 2 new 5G opportunities, one is fronthaul.

I assume that it's the high-speed (inaudible) connection.

You said calendar year second half '21 ramp.

On the second design win with your [processor] solution, I'm assuming that this is doing more of the digital front-end functionality.

When does this design win start to ramp?

And are both of these wins with your 2 existing 5G customers, or are they with different 5G customers?

Sure.

Okay.

So yes, you're in the ballpark on the fronthaul, that's the type of software we're talking about.

On the second one, it's not doing the digital front-end.

Think of it as when you go to massive MIMOs, so first of all, every radio is going to have some digital front-end processing on it.

When you go to massive MIMO, there's going to be additional processing that's required.

And so what we've seen is, one of our customers and we are not going to go into who and when on, as we talk about these different sockets from today.

But think of it as that processing needs to get added to support massive MIMO, our products are actually well suited to do that and so those are being designed in, and those are going to ramp in Q4 of this year.

So that's one that's a pleasant surprise to us.

And our next question comes from Ross Seymore with Deutsche Bank.

Matt, first, for you, on the networking side of things.

Obviously, people know you've taken the Huawei business out, and so that's on the networking side, but your guidance is for that to be relatively flat or just slightly down, I think is what you said.

So can you talk a little bit about what's going on with the core business growing?

What's driving that?

And maybe in a bigger picture sense, what sort of reaction are your other customers having if the Huawei side of the equation can't ship?

Are those leading those customers getting more eager to take share and attack the Chinese market?

Is that leading to some of the new design wins you have?

Just want to see what the competitive reaction is in a broader sense?

Okay.

Sure.

Let me take them one at a time.

So on the first one, just to go back, networking, we guided to decline in Q2.

What was said was if you put in Huawei back in, so it was apples-to-apples and networking would be up sequentially, so that's a positive.

So you should assume ex-Huawei networking is going to increase sequentially.

That's offset a little bit by the fact that some of our customers in 4G, there's been some slowdown there as pre-5G systems are shipping, and then there's some anticipation of production 5G coming online in the second half.

So there's a few moving pieces, but the net-net is that networking ex-Huawei is up.

On the second question, yes, none of the -- so the design activity we see, let me just take a high level first is, it's very robust, okay?

It's across a number of different products and technologies that we sell across a number of different customers.

None of these design wins are related to any issue with what's going on, on the Chinese customer that's going through the export restriction.

And whether or not the other competitors are going to take share, I think all those CEOs are getting those questions right now, and it's a bunch of articles you can read from those guys about how they're viewing that situation.

But just to be clear, the opportunities we see are just a result of, quite frankly, deep engagements with the customers in this space, a very compelling platform of switches, processors, basebands, all the key pieces and now, some high-speed connectivity, all the key pieces that are required to enable 5G.

So, hope that's helpful.

Yes, it is.

And so my first follow-up question, switching gears over to the storage side.

I know you just answered to a prior question that throughout the first half of the year, it's really doing what you had expected.

But you mentioned in your preamble that the enterprise side was starting to show signs of weakness, and it wasn't just a cloud dynamic anymore.

Could you give a little more color on that?

Is that just within the fiber channel side?

Is that a broader dynamic?

And how is that the same as before, if it sounds like that, that's a new development?

Sure.

So I'd say that Huawei give the commentary around fiber channel.

The market numbers I gave you, which was the 18%, that's a market number, that's a server decline and it's not all cloud.

So the commentary was really, we're seeing some of this in fiber channel, but I think it's related to a bigger picture.

And when we've unpacked sort of where the market is going, the declines are not all cloud, it's actually happening in the enterprise space as well.

And so we thought it would be appropriate to comment on that given that we've got reasonable exposure there.

It's -- and you could sort of speculate on why, but I think the bottom line is, we see that trend as being not a strong one at this point.

And our next question comes from Blayne Curtis with Barclays.

I have two.

Just -- sort of following Ross' question, because I -- maybe I missed what you said, Matt.

I'm just kind of curious what is growing?

You said, ex-Huawei would be growing in the backdrop of enterprise being softer.

Have you seen any of that softness go into the rest of your switching business?

Hey Ross, this is Jean.

So I'll add some color, right?

So when you look at it sequentially, we do see -- we have a diversified fallback lines under networking from security, switch, PHY, all those things are actually growing sequentially, which offset the Huawei impact.

And we're actually very pleased to have here with all the networking product to the classic Marvell side, the design wins continue to ramp up with different other customers across both major OEMs and the broad customers outside of China.

So that's really what's happening sequentially.

Got you.

And I do look like Ross, but it's Blayne.

The just maybe a follow up on the new design win, Matt.

I mean, obviously, the Avera deal gets you access to a lot of these ASICs designs, many of them are in the radio.

You're talking about Fusion there, kind of.

Is that temporarily?

Or is that doing some additional functionality?

And maybe there's more of a content there in the radio head, just kind of curious to your perspective?

Yes.

Just to be super clear, it's the -- the new opportunity we're speaking about is additional content, okay?

So it's independent of the solutions that you mentioned.

Now what I would say is, you think about the long-term where, if we can have a strong presence in the digital front-end then we can have the processing capability for massive MIMO, and then you can think about having all sorts of options to really leverage the technology platform and the company to optimize these designs over time.

And I think, from a supplier point of view, we are going to be viewed as extremely compelling in having all these pieces, but yes, we're excited to actually be able to, with Avera coming online, to be able to participate in that part of the market as well, which we had not previously at Marvell.

And our next question comes from Kevin Cassidy with Stifel.

I'll switch topics a little bit by asking about your Ethernet business.

Can you give us an update of what's happened with the NIC card business and also, your new Ethernet switch that was introduced, some type of design win traction you might have there?

Sure.

Yes, Kevin.

So I'll make a couple of comments.

So I think on the Ethernet mix side, that's a pretty small business for us.

You may be familiar with it.

It's primarily sells into the enterprise.

Our cloud position there is small.

So think of that as being a smaller business tied to the enterprise cycle.

For our new product, which is our 12.8-terabit switch, that's doing really well.

We introduced it about a quarter ago.

Design pipeline looks good; a lot of interest from a number of customers.

I'm not really ready to give a detailed update, but I'd say that the traction is very strong for that product, and we've been very pleased by the reception.

So I think as we -- as time goes on, we'll be happy to update the investment community about how our Ethernet story plays as we've now really entered a brand new segment that we weren't in before, addressing the higher end from 3.2, all the way to 12.8.

Right.

Yes.

That's right.

I was wondering if you even could give us an estimate of when you'd expect to start to see revenue for that Ethernet switch.

I think it's still early days, Kevin.

The design cycles depending on the customer tend to be a little bit longer here.

And so I don't think I'm quite ready to commit to that.

It's a bit early, but we'll certainly be happy to do that as we get a little bit closer and make some more progress.

And our next question comes from John Pitzer with Credit Suisse.

Our next question comes from Mark Delaney with Goldman Sachs.

The first question is, I want to better understand how the company was trying to incorporate some of the broader implications from Huawei.

And Matt, I know you said that you can incorporate the mid-single-digit direct exposure, but difficult to gauge.

And I think some of the potential knock-on effects to other customers, and for example in storage, some of your [drive] customers have some exposure.

Did you factor in any additional conservatism into your guidance in storage and network and to try and think about some of those potential factors?

Is there any other color you can give us on how Marvell is thinking about that would be helpful?

Yes.

Our current assumption is really just to take out this one particular Chinese customer's revenue going forward.

Matt mentioned it during his prepared remarks about the potential impact of RAM, how other customers they get impacted, which will come to impact us.

We are not going to be able to estimate at this point what it means, it's just something we are mindful, but it's not in our current assumption and the forecast.

Yes.

I think it's absolutely right.

And the assumption we're making is that our customers, that we're talking to, have comprehended this and their demand signals they're sending us.

And I think we're mindful of that and certainly hoping that they're doing that.

But I think it's -- as Jean pointed out, it's really difficult for us to add a judgment on top of somebody else's judgment.

And so we're -- I think the cleanest way, we just take Huawei out, it's out of the numbers, and let's hope that the whole industry is, kind of got the same methodology for accounting for this.

Okay.

That's helpful.

And my follow-up was just on the strategic road map that the company has been under way with and doing the ample of acquisitions, and appreciating all the commentary you provided along that front.

Assuming after the deal was closed you mentioned your cash balance could go up, should we think about appetite for Marvell to do any additional M&A, given the strong balance sheet with the results?

Or would you want to fully integrate Avera contrary to people thinking about any future M&A?

Sure, yes.

Let me take that part of the question.

So first of all, yes, appreciate everybody listening to the backdrop because we felt it was really important, so that part of the call was a little bit more explanatory to make sure that everybody understood the context that these weren't just sort of 3 opportunistic things that came along but rather, we were really thinking about how to drive transformation in the company during this period.

So when you think about the priorities right now, the priorities are going to be get these things closed, and then get the integration down to Avera and Aquantia, that's the priority.

I mean, certainly, we'll be in a very good position when that's done.

But Jean, maybe you want to comment on the -- I guess, you're already commented on the excess cash, but it's certainly going to be a good place to be, given the environment.

Yes.

So yes, as I mentioned earlier, over the next year, Matt mentioned that we're going to focus on integration with the 3 deals from the cash side, that we'll have a $500 million excess cash plus additional free cash flow generated by the combined business.

I think that you should think about it, our first priority that we think we have a tremendous financial flexibility to both buyback some shares and also pay down the debt.

So that will be our focus of cost in the longer term.

Our capital allocation policy has not been changed.

And our next question comes from Craig Ellis with B. Riley FBR.

Congratulations on what's been very active month of May.

Matt, the first question I had is related to your opening comments and the transactions that the company has announced.

So while neither transaction individually is anywhere near as big as Cavium, in aggregate, they're a different kind of operational challenge.

So can you spend a minute reflecting on what the company learned with Cavium that informs your view on how you will approach 3 individual transactions that all will be occurring about the same time, to ensure that you've got good execution and are driving good financials through those transactions?

Sure.

So let me address a couple of points.

So I think the first is that, you're right.

Although the sheer size of the 3 isn't the same as Cavium, I do think that in the long run, they will -- they have the potential to be a significant for the company.

So I think that's first point, that we view this is a very major step in our company's journey here.

I would actually back up the clock a little bit, even go to pre-Cavium.

When the team arrived here about 3 years ago, by the way, we, and you were early on, in the story, Craig.

We did a lot, okay?

We -- obviously, we pivoted the company around a new strategy.

We divested a number of businesses, simultaneously, in parallel, by the way, so building the consolidated design centers, tremendous amount of activity.

And what we learned through that process and what we developed was a very strong culture of execution and doing things on schedule or better and doing them on budget or better.

That bode well for when we did Cavium, and that integration, as Jean pointed out, very happy with our team.

The fact that we were able to do the entire ERP integration in about 9 months ahead of schedule, I mean, you can hear horror stories of ERP has gone wrong, our team did a fantastic job from that point of view.

And so from the operations side, what I'd say is, we've built a very strong, capable, battle-tested team here, that I have the utmost confidence in to execute.

And when discussing these opportunities with my team, everybody enthusiastically agreed that we should to go for it and was ready, and this goes from IT to business units, to sales, to legal, to HR, to finance, you name it and obviously, our operations team.

So we feel pretty good.

We know we're going to be busy, but these are extremely important for us to take our next step.

And so you think about Phase 1 as being the restructuring and repivoting of the company.

Phase 2 as integrating, emerging with Cavium, and then this next phase of growth leading into these new growth vectors.

We're pretty excited about what we've actually accomplished over the last 3 years.

And I do want to just say thank you on the call to the entire Marvell team who, quite frankly without these people, we could never have embarked on such an ambitious journey.

So wrong answer, but I think you hear the bottom line is, I'm very confident in my team and our ability to execute these transactions.

Yes, significant transition indeed.

The follow-up question is just trying to get a better sense of what you're seeing out there in the marketplace with your customers.

So it sounds like with the broad-based networking growth in the quarter, absent the effect of that large China customer that things are moving in the right direction.

But on the other hand, storage seems to be a little bit slower.

So do you think you're seeing signs of an improvement, a bottom and then an improvement in the data center market?

Or is storage really something that's driven by company-specific product performance, less so than a recovery in the demand environment?

Yes, so it's in 2 pieces, I'd say.

One is certainly a positive that X, our China customer, we do see some sequential growth, so I think that's positive.

On the storage side, it's really hard for us to call because if you think about that revenue on our controllers, it's all shipping into somebody who's then shipping into -- by the way, if you go through the whole supply chain, we're shipping into a drive company, they're shipping sometimes with distributor, who's shipping to a system integrator, so who's shipping to an end customer somewhere.

So it's really hard for us to have the sensing at the end in customer level of what's happening, relative to our other businesses.

So we just can react to kind of where we -- what the demand signals we see, based on where we are in the supply chain.

So I think there are probably people downstream from us, probably have a better point of view on that.

But certainly, if you -- high level characterization would be at least networking, showing signs of life and some movement, and then storage still under shipping what we think is the end-to-end demand.

And our next question comes from Srini Pajjuri with Macquarie.

A couple of clarifications.

Matt, first, on the, just on the -- first of all, congrats on the Wi-Fi divestiture.

And I'm just curious, if you see any other opportunities to prune the portfolio further?

It doesn't look like it, but I just wanted to ask you that question.

And then I have a follow-up.

Yes.

The answer is no.

We've done a lot here, okay, and these are some pretty big moves, including the decision to divest Wi-Fi.

So no other portfolio plans.

And then on the 5G design wins that you talked about, the 2 design wins, I think the first one is pretty clear.

I'm just curious on the second one that you mentioned.

I just want to understand the timing of the dram.

I think you said something like fourth quarter of this fiscal year.

I just want to make sure that I got that right.

Yes, you did.

You did.

That's something that's come in that's newer us, and we're obviously preparing for that.

But yes, it's actually going to be this year.

And then just to clarify that, Matt, this is on top of the baseband ramp that we've been talking about at Samsung.

And then, also, could you put some perspective in terms of, I know you mentioned $500 of additional comp in that.

What's the TAM for this particular design win, as you see it?

Yes.

Okay.

So just on the size.

I mean, the TAM is -- it's going to be hundreds of millions.

I mean, I don't have an exact number for you.

Some of this is -- this is a very dynamic market.

I think for everyone that's following us when we have these calls, even going back to our Analyst Day, I think every quarter, we've just seen the design activity and the intensity increase.

And so, as an example, this was something that the recent radio head opportunity was something that came up just within the last quarter.

So a lot of activity from that point of view.

And our next question comes from Rick Schafer with Oppenheimer.

Matt, following that 5G, I'm just trying to maybe understand it a little better, but you've spoken a lot in the past about how your 5G content is up roughly 4x versus 4G.

And I'm just curious how Avera, once it's in the fold, how does that potentially change that equation?

Is that number -- would that be more like a 5x or a 6x?

Just kind of trying to get a sense above and beyond, obviously, we talked a lot about the $500 incremental content in your opening remarks and stuff, but I just was curious if there is a rule of thumb there, post-Avera close?

Yes, I don't have an exact number for you, but I would say it's a multiple of the current SAM we have, just what is that we don't know.

But certainly, there's a number of sockets that are out there that have traditionally gone all ASIC, and we certainly think with obviously the design team that they bring, plus leveraging our technology platform, which isn't just process technology by the way, it's also Marvell's IP that could be used and leveraged to help our customers differentiate in ways that they haven't been able to before, quite frankly.

We think it's very compelling, but I don't have all the exact TAM numbers yet and I think as we -- the way I'd say it is, we need to -- I want to see how the -- this year is playing out in terms of all the things that we're working on, and then come back to you guys with a more comprehensive view because it seems like every quarter we're finding new things, and then it's -- I'd rather give a comprehensive update at a later date.

But it's...

Got it.

And then, Jean, I know you've got a lot on your plate.

But just on the Avera, I know you talked about it being a 50% or so gross margin business historically, and you talked about improving on that.

Sort of how close do you think the corporate average Avera can eventually get?

And sort of how to get there?

What are some of the bigger moving parts?

And I mean, could this be a 60% or better gross margin revenue stream over, I don't know, over a certain time line, maybe identify what that time line might be?

Yes.

It would take some time, right?

And the way to think about it is, if we're adding new design wins that we are going to get, the gross margins are going to be very similar to Marvell's corporate average.

So when you start to mix the platforms, over time, it's going to migrate into our current gross margin.

But it will take time.

Remember, all those things that have a very long cycle, typically they last 5 years or beyond sometimes.

But we are very confident over time it's going to get there.

Secondly, right, is operationally, we have a really strong operational team.

We do think once we integrate Avera into Marvell, we're going to have an opportunity to improve cost of sales and then to also help the margin improvement going forward.

So we'll provide you all details update when we close the transaction, but we do think we'll have the opportunity to migrate their gross margin to close to ours over time.

Yes, and what I'd say, Rick, too I think one of the variables that I would add on top of what Jean said is that we have to first get this closed and get the team in, and then really it sets the opportunity pipeline because in this business, it's like actually, our -- even our standard product business, I mean, the ability to command premium margins really comes with the value and the differentiation that you add.

And so, when we look at our pipeline, we also want to -- we're going to want to have some flexibility, right, because we're going to look at, okay, how much -- how many opportunities do we have that are just pure, pure ASIC?

How many are, where we could add Marvell IP, to help the customer differentiate and so that's going to have a different margin structure?

And so obviously we're a company that believes that the gross margin equals innovation, and that's what we're driving in the company.

But that being said, we also have to assess the SAM, and that's going to be a part of how that journey goes, is what opportunities we decide to take or not and what's the returns on those and the volume.

But in any case, the journey in that business will be up and to the right, both from an operational improvement point of view as well as on the new designs.

And our next question comes from Tim Arcuri with UBS.

This is John on calling in for Tim.

Adding my congratulations on the connectivity divestiture, and I mean obviously, you know our opinion over the past year on what you guys should do with that business.

Just kind of curious about in your perspective, how long were you guys looking at this move?

Which move?

The Wi-Fi.

The Wi-Fi.

Well, let's be clear.

The -- we were approached on this business, right?

We didn't put it up for sale.

We were approached on this business, and we were approached by a number of parties, and we elected to have those discussions, and then it turned out to be a great outcome.

But that was really the catalyst, was just a number of -- a number of inbounds that looked compelling, and we felt it was our obligation to take those seriously.

On our own, we were driving that business, right?

I mean, if you -- I mean, you know, we were -- that business 3 years ago was low to mid 30% gross margin that had been declining in revenue.

It had all kinds of issues.

And credit to the team here really improved that business structurally, re-pivoted it.

Got their mojo back, and the business has really got to a very good place.

And that's why you can look at the valuation we got for it because there's a lot of value that's been created.

And I think by the way, I think NXP will do a fabulous job with the asset.

So -- but yes, we -- you can see from what was paid for it, it's a good business, and we were fully prepared to drive it on our own as well.

Okay.

All right.

Fair enough.

I appreciate the color.

My follow-up, in terms of your storage business, I mean, can you give a little bit more color in terms of SSD versus HDD dynamics?

And what you see for the coming quarter as well?

No.

We don't provide that comment to details, right, because we also have fiber channel in that business and some of the other product lines.

So the way we look at it is actually from, on the market perspective, we look at the client, we look at the enterprise data center and periodically, we'll provide you with the details of how we are migrating to increasingly address the enterprise and the data center market.

That's how we look at it going forward.

And our next question comes from Gary Mobley with Wells Fargo Securities.

In the interest of time, I'll post both my questions at the same time.

In the just concluded April quarter, how much buy ahead was there from Huawei anticipation of the shipment ban?

And Jean, you had mentioned in the past that you can bring the non-GAAP OpEx down to perhaps $280 million, and I think you laid out a time frame of maybe the fourth quarter of fiscal year 2020, and you keep under promising and over delivering on that front.

So I'm curious, in light of the revenue headwind from Huawei and other factors as well, should we be thinking about maybe something better than $280 million by the end of this fiscal year?

Yes.

So firstly, on the Huawei question, right, the ban happened 2 weeks into our quarter, so we did not ship too much to them during the first 2 weeks of the quarter.

And then, if you look back into the last few quarters, it's very difficult for us to tell how much inventory they are building, but our assessment is we believe the run rate business of Huawei is around mid-single digits as members of our revenue.

So that's how we think about it.

Secondly, on the operating expense question, I think that we continue to target asset in Q4 fiscal '20.

We'll get to close to $280 million operating expense, that's how we are targeting, and I think our team has done a great job to manage OpEx each quarter.

We'll continue to be disciplined in managing the OpEx to get to our target level.

And our final question comes from Christopher Rolland with Susquehanna International Group.

Guys, 2 quick ones from me.

I guess, first of all, given the Huawei ban, do you think this going to help other OEMs guys that you are more exposed to and hence help you guys?

Sure.

Yes, Chris.

So on that one, again, I think there's quite a bit of commentary on this that's out there now.

I think Ericcson's has been asked, Nokia has been asked, Samsung's been asked, you go through to the list.

And so you can sort of listen to their view.

I mean, I think what they'll tell you is they're short-term positive, there's also potentially long-term negative side.

It's hard to call and it's also hard to call when the situation resolves itself.

So we don't have a particular point of view other than the customers that we are engaged with are doing well on the market, certainly on their tenders and their bids and their positioning for the deployments coming up, and the design activity with those customers is very, very high.

So that's the commentary I would give you, it's hard for me to handicap the global base station market, puts and takes right now.

Great.

And then lastly, have you guys, in the industry, heard anything about China withholding kind of rare earth materials?

And if that would have any effect on the hard disk drive market?

Or do you think that is kind of a low risk, low probability situation?

Yes.

Don't know, not the expert on that, although I'm reading the newspaper over the weekend kind of thing.

So I don't have a particular view on rare metals and then how it's going to trickle into the supply chain at this juncture.

Thank you.

Ladies and gentlemen, this concludes our question-and-answer session for today's call.

I would now like to turn the call back over to Ashish Saran for any further remarks.

Thank you, everyone, for joining us today.

And we look forward to seeing you at upcoming conferences.

Thanks, and goodbye.

Ladies and gentlemen, thank you for participating in today's conference.

This concludes today's program, and you may all disconnect.

Everyone, have a wonderful day.