Seagate Technology PLC (STX) CEO Dave Mosley Presents at Morgan Stanley Technology, Media & Telecom Conference (Transcript)

|
About: Seagate Technology plc (STX)
by: SA Transcripts
Subscribers Only
Earning Call Audio

Seagate Technology PLC (NASDAQ:STX) Presents at Morgan Stanley Technology, Media & Telecom Conference February 26, 2019 2:00 PM ET

Company Participants

Dave Mosley - CEO

Conference Call Participants

Katy Huberty - Morgan Stanley

Katy Huberty

Let's go ahead and get started. Good morning. Thank you for joining us. I'm really pleased to welcome Dave Mosley, CEO of Seagate. He took over the CEO role in 2017, but has spent 20-plus years running really every aspect of operations at Seagate. So we're really pleased to have him today and look forward to the discussion.

Before we start, please note that all important disclosures, including personal holding disclosures and Morgan Stanley disclosures, appear on the Morgan Stanley public website or at the registration desk. And then, I think Dave has to make a quick comment as well.

Dave Mosley

Yes, our risk factors because we are going to make any forward-looking statements that are risk factors and uncertainties that are filed with the SEC and on www.seagate.com.

Question-and-Answer Session

A - Katy Huberty

Dave, thanks again for joining us. I want to start by talking about demand drivers. Seagate has partnered with IDC over the last couple of years to start framing the impact of new workloads like artificial intelligence and automation and Internet of Things with a framework that suggests data growth to be 27% annual growth. I frankly think that number seems even a little low, but when you put that in context historically NAND flash and HDD capacity growth in the teens, there is a really bullish potential takeaway. Give us your thoughts as to how much of that data and what types of data you think will reside on hard disk drives relative to NAND? And does that bullish outlook suggest that at some point the hard disk drive industry might actually have to add capacity to serve that demand?

Dave Mosley

It’s a good question, really hard question to forecast. So all of us believe the data is growing and the trends that we should be looking at are forward trends not reverse trends. So if I think about the past client server days, a lot of the drives that we shift were underutilized, not utilized at all the capacity wasn't utilized. That’s not true today in mobile cloud. There is most drives are working 24x7 in the data center. They are full. Their access is very aggressive. Workloads are even stronger than what we used to mission critical back in the day.

And if you think about the next 4 to 10 years, all this growth is going to come. I think the one thing that we are certain of it is that, it's not going to be all mobile cloud. There is going to be a number of different things that we call the edge that really get into latency and cost optimization. That may look like data center, some of them they look more like PCs, but there are a myriad of different applications coming.

What technology goes into each node is probably -- that's probably still going to be a panoply of different accessing technologies, but I do think that there is plenty of hard drives, I also think the IoT for example has plenty of flash. I think just like -- I would recommend run good factories and make investments wisely. I think there is going to be plenty of demand out there at the top of these cycles for everyone.

And I do think we are going to have had capacity, we are going to be losing capacity in components much the same the silicon does because it becomes more and more complex to meet the kind of design features and cost abstracts that we have to meet in the future. We will have to answer that call inside of our factories with more and more complex designs that will mean we naturally lose a little bit capacity, and we are going to have to add capacity to hit this upswing, but it’s a tough transfer function, it really is.

And I wish it was easy, but it's not. Who are the customers going to be in five years? What kind of service levels will they want? These are all big questions for us.

Katy Huberty

You made a really interesting point about that client server to cloud workloads transition because in the clients server world, there's a lot of short stroking, significant underutilization of drives, in some cases, 20% and that transitioned from client server to cloud, it's short of now played out. And you know the market arguably sees the real level of data growth in a way that it has been over the last couple of years.

Dave Mosley

I also think that the futures are the IoT, that's going to be a world where there is a lot of writing and a lot of reading going on rather than just write once and read many or write seldom, read never you know kind of some of the old applications that existed before, and I think the edge is definitely like that too. If you point the surveillance unit, for example, we have you know 64 cameras going to 128 cameras with 9,000 hours of HD video, all pointed at one or two hard drives. I mean these are really, really rugged applications. The workloads are pretty stringent as much as mission critical was in the old days. It's interesting to see what some of these futures will hold.

Katy Huberty

If you look at recent data from IDC Seagate gain about 300 points of revenue share last year, help us understand you know is that cyclical or is that a structural gain of market share? What do you think the drivers are? And then, how do you think about balancing market share and margins going forward?

Dave Mosley

Yes, I think the first thing I'd say is market share is not the sale for us, if you remember three years ago I think, we lost a fair amount of market share. And what we were doing at the time was we were moving our capacity points up, out of markets that maybe cannibalized at some point, so that we have a little bit more -- so we were the prime movers up into the higher capacity, it was a higher performance level. I think that serves us well. We have a fairly stable portfolio across all products. We opened the way we serve customers, but some of them have very diverse portfolio, some of them need a legacy products still from us, some of them need the cutting edge, the newest, the highest capacity, the highest performance going. And we can't just go in there and pick and choose all those things ourselves.

So I like our portfolio the way it is right now. For some reason, our industry tends to focus on only the highest capacity points, but that's actually a fairly small number of units. I think if we need to keep, see the move to market up in capacity points we can do that. We were staging our internal components as a disk for that as well. And the market share gains I think are really a function of how we serve our customers and how strong our portfolio is and rather than a strategy or something.

Katy Huberty

Okay, Samsung has recently talked about 20% declines this quarter sequentially in NAND prices, last quarter 15% decline in NAND prices, and the concern of investors is that cheaper NAND means acceleration of cannibalization of HDD. Is that something that you would expect to see in the business over the next couple of quarters? And how would you describe Seagate's exposure to those areas of HDDs that are being cannibalized?

Dave Mosley

Yes, we said publicly for example for the last two years I think 15K RPM and 500 gigabyte notebook driving, and we've lumped those into percentage of our revenue that we thought was ripe for cannibalization, I mean it's still 6% of our revenue and there's still customers who say, I want those products, I need those as part of my portfolio, please continue to make them for us, so we will. We're not investing R&D that we're rather in harvest mode on those technologies. You know I think the tactics of dramatic price increases or decreases have affected the market.

There is a dramatic price increase a couple of years ago and then the dramatic price decreases. -The market is -- doesn’t respond that quickly though because you think about our customers and their marketing messaging and their architecture that they have. Sometimes, they need the legacy products so the market will smooth some of that out. It will have a cannibalization affected and it's been going on for years. And we're factoring that into our go forward claims and how it points our R&D dollars and fundamental head supply chain and disk supply chain, so that we can service the right markets two to three years from now.

Katy Huberty

And do you think that 15K enterprise and the 500 gig into low network or are those still but by categories that are most of risk some investors push back and say, well now it's going to become 1 terabyte.

Dave Mosley

Yes, it's probably will, so and there is some of that depends on what height of submarket you are talking about PC commercial just PC in general, but then there is other things that are not very susceptible. For example, gaming computers are all dual drive systems now and I should say all, but tend to be dual drive system of high capacity SSDs and HDDs. So I think it's too easy to just say, this entire market will go, will tip quickly. We have our own forecast for those things obviously and we're trying to make sure that our dollars are pointed to the ones that are going to continue to grow and not the ones that are in harvest ones.

Katy Huberty

But certainly at this conference and over the last month or so, there is a growing message out as, component suppliers with the cloud cycle that we're in is broadening and so that will be a recovery in the back half of calendar 2019 and you've messaged that as well. What drives your confidence so that's the right comment?

Dave Mosley

Yes, I would say this is predictable as I want. This every cycle is different and mostly the cycle that we've seen before are I'd say supply driven, but either by pricing increases like we saw a couple of years ago or availability. And this one actually seems differently, this is a little bit demand driven, and there are lot of people who can speak eloquently about that, about what's the IOs are doing and why they are making the decisions to move the cloud and whether the cloud is really ready for them and whether it's things are going to be repatriated. Those discussions are really interesting discussions, but all I can do is talk to the customers and ask, what do -- how you think this is going to turn back on and what kind of capacity points you need.

That's why we made reference in our earnings call about some of the fast three point transitions that are happening. When the hard drive raises its capacity from say 12 to 16 gigabyte or some terabytes in the one datacenter application, that's a huge TCO proposition, and if you wait a few quarters for that, it changes the dynamic so how you look at that install which you are going to do for five years or seven years or how longer that ever long you are planning to use it. So that can cause a little bit of trepidation in the market as well.

And so I think the start of the cycle if you will was demand driven, but the how fast you come out of the cycle and really some time to how fast can we answer the call. And then, the dynamics are going on with the cloud service providers and more broadly speaking, geographically, all the people doing cloud build out, I do think that there are clouds build-outs starting and there are certainly buildings the built, and that is certainly year that's in the datacenter that exists in the world today that are aging.

They need to be replaced and there is a lot of reasons you've replaced them. Some are powers, some of them are for just pure age, you can upgrade the capacity point something like that. So I think this kind of visibility. It's going to turn back at some point, I don’t we can't predict exactly when.

Katy Huberty

From a technology roadmap perspective, Seagate and WD have somewhat diverging technology roadmaps with HAMR and MAMR technology. Talk about HAMR for Seagate, the advantage that that brings to the customers? And why you think it’s a better roadmap than MAMR?

Dave Mosley

Here is what I'll say, it's not really all that different, so you are talking about the way that we apply assisted magnetic recording, the way we apply more energy into the interface, such that we can flip the media BIPS. Actually the media BIPS are -- actually -- the individual grains are probably more important than the discussion media is than the writing process.

So, smaller grains gets you higher densities and higher densities is what everybody wants to continue to build out bigger and bigger data centers with higher capacity points. So when we want 20 and 40 and 100 terabyte drive some day, we have to get those grain sizes down. We actually have the technology to do that in the industry, but you need to be able to write it and that’s why -- that’s where the assisted magnetic recording comes in.

I think MAMR could be a good technology. We worked on it. I remember I have said this before I remember in 2007 we were really excited about it. So it's not going to work, it's just what we feel is that we need to get onto the HAMR transition because we want to get these smaller and smaller grain sizes in the media so we can get the higher capacity points. We have been working on it for quite a while and we are pretty happy with our progress. We will ship it when we can make things working high volume for our customers.

Katy Huberty

You have lived through the number of technology cycles that as a company how were the HAMR adaption cycle compared to something like PMR and other technology NANDs?

Dave Mosley

From the customers' standpoint, this is actually interesting because we get into these discussions a lot in the industry, but customer standpoint is it's the disk drive with higher capacity point and can identify more of them. I think it will come down to cost, performance, reliability, power. I think we can solve all those problems and make people really happy to go to 20 terabyte, 24 terabyte, 36 terabyte so on. The technology exist in the lab, it's a matter of turning, getting onto that curve is really the hard part so getting the first foothold. We believe we are there from the reliability perspective, but still have to make 24 terabyte, 36 terabyte areal of density points work in the drive and those aren’t just challenges about heads, those are challenges about heads and BIPS and tracking systems and electronics and power and everything else.

Katy Huberty

Do you have customers today with 15 terabyte HAMR drives that they are testing?

Dave Mosley

Yes.

Katy Huberty

What's been the feedback?

Dave Mosley

All the feedbacks have been good. I would say the general category it looks like a normal drive so it looks like the other drives that are out there. We will ramp it when we are ready to ramp it and there are some big problems that we still have yet to solve that the teams are working very hard on that. I think in the scheme of all the problems that we solved to get through 16 terabyte drive on conventional technology, the problems that we have from here on to productization are not insurmountable. They're actually fairly straight forward, but we got to get through them.

Katy Huberty

And how should we think about the financial impact to the Company as when you decide to ramp HAMR aggressively the implications for margins and costs?

Dave Mosley

There are small cost impacts as you start to ramp and I think this is one of the areas where there is a lot of FUD being thrown around. I think some of those things actually are easy get -- relatively easy to get through. And then further out, I think we can continue to work those costs down. Any new technology point you're going to probably have to add more process content in your wafer fab anyway and your media fab anyway. So you’re probably thinking cost impact just to get some higher technology point anyway is HAMR worse or better than any of the other options that you have, it's probably about the same. And I think you know the way I think about it over time, it’s our job to continue to work that, provide the value that the customers and not have them today talk penalty but rather see the dollars per terabyte continue to go down.

Katy Huberty

Those 15 terabytes you'll have the PMR drive and a HAMR drive in the market. What's generally the price premium for the new technology? And what's capacity point the PMR technology not work anymore and you're sort of forced to shift over to HAMR.

Dave Mosley

Yes, we're really on the cusp of it, some of these discussions are a little bit funny because we talk about 15 and 18, if you take it to the extreme what if we were talking about 36 versus 38. I mean that extra 2 terabyte you know, we're not going to be able to charge this price premium to that or you just want it adopted. So that's one of the reasons why I think ultimately HAMR has to get to 24 and 32 and 36 and these kind of figure points and PMR just came there. Those options aren't accessible.

Katy Huberty

Last year, the calendar year, the enterprise performance of you know not the cloud business but the traditional mission critical applications, capacity grew 30%, and this like we talked about is an area of the market that investors have been worried about structural decline because of cannibalizations from NAND. What do you think drove that capacity growth in 2018? And should we expect that there can be growth again going forward?

Dave Mosley

It's a really interesting question. I don't know how to get through it quickly. Let me try. People have been talking about the demise of mission critical for a long time, but mission critical was never a simple one size fits all application space. So, I think for example a year ago, we lost a lot of 300 gigabyte drives, which I would call boot drives and we lost those to NAND. And instead, ourselves and one other competitor did the-- finished off the last generation of 10K and 15K products which are capable of 1 or 2 terabytes you know kind of capacity points. The market has moved to 1 or 2 terabytes. Now, we lost all the 300 gigabytes. We lost all that unit share if you will, but that was never really a tractable point anyway. If you look at the systems themselves, we had systems where some of those 300 gigabyte mission critical drives were actually boot drives in all flash arrays, it's crazy as it sounds.

Katy Huberty

How interesting.

Dave Mosley

Or in the cloud or they redeemed use of boot drive, now these drives that we’re talking about the 10K drives in particular the 2.4 terabytes, those are going in packed. So those tend to playing in systems that have 12 or 24 drives in them probably have memories front in, DRAM or NAND probably both. So think of them as many cloud architectures and they're solving all kinds of different creative interesting workload. And so, I think that's a fairly stable market for a while. What's very interesting in that, it that's when you get into this concept of actuator density, so the idea that the discharge maybe gets faster from an access perspective, from a data rate perspective.

And then the more actuators that you have, so you can share the workload against those actuators with a higher capacity which you have, with the higher performance you have. There're a lot of requirements like that from the big cloud all the way down. So, that's why we're developing mobile actuator drives. So the way I would think about it is, some of these applications that are more attractive but when mission critical, long-term will become the multi-actuator big drive types of application.

Katy Huberty

Okay, understood. WD has said that they are shutting down production for that enterprise performance segments if their business is down 50% in recent months. Should we think about Seagate absorbing that market share and how much is left over the next couple of quarters?

Dave Mosley

Yes, and I think that's what one of the things it's happened is, ourselves and another competitor continued to do the -- invest. We aren't anymore, so we are not doing anymore 10K or 15K, but this is the same trajectory now. We finished up the last generation about two years ago and we will continue to harvest that as long as they want to stay around. I think it could be around to quite some time.

Katy Huberty

But a year ago you entered into a long-term NAND supply agreement with Toshiba. What are the advantages in your mind to partnering versus vertically integrated owing a NAND business?

Dave Mosley

This gets into some of the battle starts as the last few years, as we were trying to develop our SSD business, we believe that we have the good customer relationship to build that an enterprise SSD portfolio. We believe we have the controller technology, we believe we have the firmware technology and the knowledge of the customer system. But probably our biggest problem was the transitions from vendor-to-vendor and NAND node to NAND node inside those vendors. So as NAND went from 2D to 3D, it is really hard to keep up.

The good news is that customers didn't really blame us because of what's happening industry wide, but there are just still many changes for the next cost point or because the liability whether it's wanted to be on one node. I think that industry dynamics settling out a little bit, but having visibility and a partnership, Toshiba has been a great supplier to us for many, many years. So having this supply arrangement with them is big in our ability to not only to be guaranteed but we actually have access to the capacity that we need, but also that we can predict when in the next node change is happening or understand what the defect structure is of those nodes we have in front of time and we can tune our solutions accordingly.

Katy Huberty

So your goal is to clearly grow the assets key business overtime, but in the current quarter you've got about business down sequentially as we go through some products transitions. How soon should investors think that SSDs can return to grow some long-term? How would you describe the Company's strategy about how aggressively you'll pace that market?

Dave Mosley

So, it's a pretty aggressive thing. I think in Q4 last year about double-digit growth quarter-over-quarter all through this year or so, and we did that for the first three quarters. This quarter that we are in right now, there has been a lot of dynamics going on in the industry. So we have a back up that I think that's what you're referring to. I think that's a short-term temporary thing. Our customers like us. We don’t have to change technology nodes. We don’t have to change controllers to go, to go get more business when they come back. Our customer base is still fairly tight. We can expand though. We have the portfolio to do that. We are talking about SaaS and NVMe solutions. So I'm pretty bullish on this in the future. I think we can find home for all those bits with those customers.

Katy Huberty

Within the non-HD business includes SSDs and then there is a systems business and there has been a strategy to move away from the low margin part of the systems business. How much work you have left to do? And what could be the impact of gross margin as you exit the margins?

Dave Mosley

Yes, I would say we are basically done with the heavy lifting. When you think about that, we acquired the businesses that I like them but they are very focused on serving client server customers. And what that tends to be is -- it's an oversimplication, but they tend to be competing with ODM, during configuration, with a very complex configurations and very low margins and things like that. And of course that’s not what we wanted to do. And once you have all these obligation to these customers you have to continue to need those obligations, but you really like to try to steer towards a lot less configuration complexity and a lot richer Seagate content.

So I would much rather be selling systems who have lots of Seagate drives in that and the take of the -- the companies that we acquired they were many times where they had no Seagate content. So, we had to transition from that portfolio to the future portfolio. The future portfolio I think does not want the complexity of the past. We want the simplification and the cost and service capabilities that come along with it.

And I think we are through that transition already. So, we are at a point where, if you want to buy a Seagate drive you can buy in the new systems and there is still some customer complexity a little bit. But I think if you have five years from now, people don’t want that complexity as more they will be able to -- we will be able to contain it into some of these systems a little bit what business wants.

Katy Huberty

And so of the 800 million roughly non-HDD revenue, now how much is SSD versus systems? And will that mix remain relatively stable going forward?

Dave Mosley

Actually, I think both of them will grow. We haven’t broken out how much is each, but I think both of them they are growing. And I think over the next couple of years we can do that well over a $1 billion. So, we will focus on growing it profitability to your earlier point about profitable. Those earlier models weren't -- didn’t have a lot of profit in them because of who you are trying to compete with. Now, I think we can try to grow this profitably and do right things for our customers and that revenue we help your revenue.

Katy Huberty

The PC market has been impacted by CPU shortages. How -- what is the impact to your PC units? And when would you expect that headwind to stay?

Dave Mosley

Yes, this is one where my perspective is little bit different than somewhere stuff I have heard out in the press and whatnot. From my perspective, there were a lot of mix changes of the big customers because of the shortages that were happening in the market. And big customers know how to ride through some of those. It's painful for them, but what I have been my mix shift is usually the kind of shift into either stuff that they have or stuff that they have that takes them into the future with the chips that they are allocated.

Some of the smaller players have -- and I'm not talking about smaller OEMs, I'm talking about people in the channel that they can get parts, they're in trouble. And they will end up in a different dynamic where they actually have to sell the bunch of products just to generate revenue because they couldn’t get the exact price that they needed. And I think that’s been the bigger disruption. And I think we can see that the distribution of channels and how it's disrupted. Some of those people tend to be become more innovative, it's not just like box PC. I think that’s old.

I think somebody is going to either servicing very distinct market segments, think bars, or they tend to be some of the more innovative people doing next generation build out. So I think there is a -- and it's not just because of the chip shortages it's all the time the component changes that is happened. There's credit liquidity issues in places the world that has impacted all this, but I think that's caused the market to be very tentative in the first half of this year. Eventually, people will settle up when they're normalizing and finding new plans or some of them probably won't unfortunately, but most of them will and then they'll say, again, the good time they need to be predictable and everybody will get back on the horse.

Katy Huberty

Surveillance is a business that's included in your non-compute segment and has grown very fast. There's a sense of lumpiness to them, inventory build and depletion. But what do you think about the demand momentum in that segment going forward? And what do you think the normalized growth rates are?

Dave Mosley

It's hard to continue to grow at 35% CAGR like surveillance has been and that's an exabyte growth. Unit growth has been a little choppy, but if you think about what surveillance is, it's an edge video cache and even DVRs back in the day were edge video caches, they weren't used as aggressively as surveillance boxes because surveillance boxes tend to have a lot of IoT devices around them. I actually think that this whole market is going to morph into maybe something that's not called surveillance any more, it's smart cities or things like that but it really is more about the edge and how you're absorbing on this data coming from the IoT.

So, I think the capacity growth is still going to be huge, there could be unit growth as well, and the hardest thing for us is the requirements that are coming down onto our system. It's not a PC drive anymore. It's got so many different streams coming at it and streams that have the services at the same time it’s a lot like what's going on in the enterprise with multi-actuators and very complex firmware algorithms that have to manage all this. It also the lowest power and lowest cost will work well.

You know, it's not going to be 35% growth, it'll probably be the sub-20, but that's after the base what it is today, but it's still going to grow very, very healthy I think in the future.

Katy Huberty

And you talk about those drives much like nearline or cloud drives have lots of heads and depth. How do the margins compare to that?

Dave Mosley

That's right, that's exactly how we think about it. And from our perspective, if we lose a head drive that with a PC maybe in a factory, it has two heads and one disk, but you're putting in the factory and it's a factory infrastructure a drive that's working a lot harder and has a lot more heads and disks. We should be able to pivot you know and make sure our investment we get paid for us so that we can make it the way through the pivot. So I think that ultimately that puts us in a better state.

Katy Huberty

Given some of the macro impacts you just talked about in the channel and where we are in cloud cycle. You've guided the gross margins down in the March quarter to lower your production and essentially not overbuild in a period of demand weakness. But you guided the gross margins of 26% which was 300 basis points above where gross margins troughed back in 2016. Do you think the industry structure and Seagate's competitive product line up is allowing for more stable margin cycles than we've seen in the past?

Dave Mosley

Well, you know maybe back to the earlier point, tell me how deep and how wide the trough is and then I can be more predictive and that's the hard part right now with our customers, But our footprint is certainly a lot better than it was two or three years ago because we took factories offline, we reduced a lot of our fixed cost to be able to handle these things. I mean we've been around long enough to know that the growth is not going to be 5% per year for the next 10 years. It's going to be very-very cyclical. Sometimes the cycle has something to do with our technology investments to the earlier point, sometimes they don't. And so you have to be able to the pull hand break.

I like the footprint that we have right now. I wish that we would have predicted this one a little bit better because I think we can get out in front of this, two quarters ago, better. But when you are in, you -- my opinion is you don’t build the wrong stuff and then have to move that later. You take the medicine, when I just said on my earnings call and then you move on. We think in the back half of the this year of course the demand and the cloud is going to come back and I think some of this other dynamics I was talking about with the channel and I think that does that does it should be healthier. So we hope where at the bottom of the trough right now.

Katy Huberty

And there is cloud the biggest driver to get back into the long-term 29 to 30.

Dave Mosley

Yes, certainly, that with most of our investments seeing in, some of them in the components through the heads and disks. That would help us to get out. And the reality is how we're going to have to build those within inside of our factories and we're going to build those heads and disks in anticipation to the demand that we will take advantage of when it comes. So, that's why we have to stay very connected with our customers. It's good this dialogue has happened about exactly how we are going to go through this transition, how many will you need in the back half of this year so on because the lead times are getting long as well. That will help us absorb our factory spend, if you will.

Katy Huberty

Over the last several years, you were able to reduce that back by 130 to 200 million each year. Are there still opportunities to take out cost? And you now have the cost structure aligned given the best you made in HAMR and in new markets?

Dave Mosley

Yes, it's quite painful to do that in the last couple of years. The most recent one that we just said was more of a voluntary retirements, and we were pretty hoping about that's good to up to those kinds of options as well. I like the footprint that we have right now and it's part of the redundant, but that's how deep and how wide with the trough that will help us. I like the -- I also like the investments we're making for the next 3 to 5 years in R&D, CapEx, and how we have to run the tools and things like that. So I like the footprint. We can always make more cuts if we had to, but it maybe to expense something we come to the couple of years. So period in the next year or two, if we made those cuts we might look good, but we are still trying to look good, making sure we stage ourselves for the ultimate growth that we've been talking about.

Katy Huberty

I'll ask one more question and we will see if there is anyone in the audience. We talked about the systems business and SSD business. It's some of the competitors in the industry have chosen to expanding those segments through M&A. How did you think about internal R&D versus partnership versus M&A moving into some of those adjacencies?

Dave Mosley

I don’t really have a preference for internal versus M&A. I mean we've had a couple of experience where the M&A has not done as well as we wanted it to. And I think there is a lot of reasons for that when you talk about stapling something on to Seagate specially maybe the business model is different or the customers serve a different or it's like I talk about before, it's a little bit of legacy. The capability you have but it's legacy you really want to be poise for the future, all the probably investors we make 3 to 5 years so I have to think about the returns on 3 to 5 years, putting that on top of the Seagate.

I think we will get into a mode eventually where there is more M&A right now. We were obviously focused on that 3 to 5 years with our own OpEx spend. I don't think it's a -- I don’t think I have a strong preference is to which way to do it, but I would say that there are things to be careful of it, it's not easy necessarily to just bolt something on to Seagate and do the and purge the SG&A and then adopt that into -- that's the problem we've had. And I think we need to think about ourselves with 3 to 5 years before we could actually see that growth or accretion, if you will to our model with those.

There are new technologies that are going to be required to answer some of the growth of the edge. And I think we will even have to make those investments internally or externally and so we'll keep looking at that.

Katy Huberty

So let me stop and see, if any questions in the audience. One down here.

Unidentified Analyst

I think IBM recently talked about thick drives, if coming back particularly in the cloud, replacing some nearline that you could use tape in conjunction with NAND. The NAND was sort of first couple of minutes of a video and then you could in that time access the tape and you wouldn’t need the nearline drive. Is that a risk? I mean tape is cheaper than nearlines.

Dave Mosley

Yes, I don’t think it’s a risk. I would say this if you look at the IDC report that Katy made referenced to earlier, there is the growth of tape. I think that presupposes from technology investments will have to be made and there is also some used cases that have to be understood much better. I personally think that nearline drives are going to -- when we talk about giving the 20 terabyte or 40 terabyte and beyond, I think they are going to provide a lot of value to be able to give the data back very quickly.

Tape is a great technology that’s been around for a long time, but I don’t think it's going to grow in the growth mode, if I would do implication. And I don’t think it’s a risk, so it's anything, probably people would say kind of data is going to matter more and you can only analyze the data sets that you have them relatively available. So, unfortunately, tape is way too cold. And it requires a lot of manual intervention to go manage or things like that so on.

Katy Huberty

Also on aren’t that many tape providers I mean there would be -- there is one right. So there is some risk if you are a cloud provider. You are always depending on supplier. Question back here?

Unidentified Analyst

If you just elaborate more on the cloud CapEx environment? Intel is just in here and sounded pretty confident in the back half recoveries. I'm curious if you're either seeing any indications or have some historical precisions around timing that to give you any sort of visibility?

Dave Mosley

Yes, we are seeing indications. The people are talking about at least. So, the way I think about it I answered this a little bit early, but I'll try to be little more clear. So, they are building new buildings that are new data centers build-out, if you will. And then there is the replacement of the old and a lot of times people do that for power. For example on servers if you can replace a 100 servers with the one new server and the powers lower and the maintenance got to lower and things like that and real estate footprint lower even do that, right.

So, there is a lot of those dynamics that play in as you have those conversations with the customers, we see indications that there is an appetite there as well. Yes, it does get into people's publicly stated CapEx. Unfortunately, they don’t break that down for me and tell me how many hard drive that mean until the last minute, but that’s just kind of the nature of the game.

The other thing I do think is the geographically worldwide, everything shutdown at the same time and I think different model, it's not just the cloud service providers in the U.S. Different models worldwide will start opening up and that’s why we have some confidence through the back half of this year.

Katy Huberty

You've mentioned earlier that some cloud providers maybe leading to a 16 terabyte because of the TCO advantage. What's the timing in terms of when you can deliver that at scale and free up some of the spend?

Dave Mosley

Yes, we said recently, we would be up early in the first half of this year, calendar year, so it's imminent.

Katy Huberty

Is there another question back here?

Unidentified Analyst

Just on gross margin and recovery, I think you've talked about one or two quarters get back into your target sort of gross margin range. You know if you think about, I guess looking into June, do you get there just getting back to shipping any consumption from March to June? And then, how much is really incumbent upon the cloud business coming back to be in that range?

Dave Mosley

Yes, there're some assumptions that do require the cloud to come back a little bit, so that gets into the mediacy of the stop that's happening in our fiscal Q2 and then translates into Q3 as well. So some of it’s the cloud coming back, some of it's the channels turning back on as well. And then how do we manage our factory this quarter and next quarter, really those are the three big variables since how do we get back into our range. We're definitely focused on how do we do those as quickly as possible.

Katy Huberty

There's another question in the back.

Unidentified Analyst

Yes, I was just wondering if you could look out call it 3 or 5 years. What would you think from a percentage of revenue would be a nearline or in the hyper scale public cloud providers versus the edge?

Dave Mosley

Yes, it's a really interesting question because it comes down to how do we define the edge, because some of the edge definition is really private cloud like, but I could very easily see nearline itself going to over 50% of the revenue and I probably biased more toward those public cloud data center build outs would be, I would call that the cloud, if you were there. And then the edge would be more like local edges, retail edges or small data center installations, cell towers, things that are -- thing are going to be serving data out to more branches if you will.

But I could definitely see it's going over 50% in 3 to 5 years from now. I think the other interesting thing for us is where the customer's going to be, and that's not necessarily that clear. I mean some of our existing customers who want to try to move there, but how's actually that moves going to happen and then there maybe new customers as well, so we have to stay focused on that.

Katy Huberty

When you say that the question was, who will the customers be, is there a kind of figure out will it be the manufacturing company and the automobile maker versus a middleman who provides a data center that's sitting underneath the cell towers?

Dave Mosley

Yes, yes, and yes. The middlemen that actually don't exist in those spaces today, it could be big players, but are already big players in infrastructure. So I think there's a lot of people with aspirations and how that's all going to play out is tough. We can talk to a lot of people and they can say I have a model we’ll start stuff and the year later they say, I don't like that model, I'm changing that model. And that’s the big problem we have in the edge data. There's no doubt that it's going to grow there.

Katy Huberty

Okay, any last questions? Guess, we’ll wrap it up, thank you so much for your time.

Dave Mosley

Thank you.