Micron Technology's (MU) Management Presents at Nasdaq 36th Investor Program Conference (Transcript) – Seeking Alpha


Micron Technology Inc. (NASDAQ:MU)

Nasdaq 36th Investor Program

June 15, 2017 07:00 ET

Executives

Ernie Maddock – Chief Financial Officer

Analysts

Unidentified Company Representative

Moving on with the program we are very, very pleased to welcome Micron from Boise, Idaho. Mark [ph] was getting miked up, so I will leave it to Mark. Okay. Ernie, welcome.

Ernie Maddock

Thank you.

Unidentified Company Representative

Thanks a lot for joining us. So I think we’ll start off with the Micron fireside chat, so today we are very excited to have the CFO, Ernie Maddock, he joined in 2015 as the CFO. He was also previously CFO of companies like Riverbed and Lam Research, and also serves on the Board of Directors of Intersil Corporation. So, thanks for joining us today.

Ernie Maddock

My pleasure.

Unidentified Company Representative

And I think — maybe, if it’s okay, if you could just start off with — you guys have a new — Sanjay has taken over recently from Mark Durcan on — in May. What’s the message about the change and what should investors think about it?

Ernie Maddock

Well, Sanjay has been on the job now about five and a half week. So it’s — I don’t think there are any big strategic changes that he has thought about thus far but he is taking I think the appropriate amount of time to learn and understand the business, of course he is intimately familiar with the NAND business. He has not had experience with the DRAM business before but you know, he is coming upto speed there at a lightening pace and his joining was coincident with our fiscal year which starts again in the early part of September, so we’re in the middle of finalizing our annual operating plan, so this is actually really opportune time for him to understand comprehensively about the business.

So I think it’s a little too soon to tell whether or not there will be any major strategic changes, it feels a little less likely than a little more likely at this point but certainly I think that he will help us achieve really outstanding results relative to both pieces of our business given my experience so far.

Unidentified Company Representative

You know, I want to ask you a few questions about just general industry trends because I think what’s happened in the memory industry is fascinating. If you look at the NAND market for example; you know, if you look peak to trough, 15/20 years ago you would think about 50% peak to trough reduction in revenues and that’s causing big kind of dislocations in the cost structure and right now if you look last decade, that peak to trough kind of went from — to 20% from 50% and then more recently the peak to trough was 30%. So it seems like a lot of beta has come out even as the industry has become more profitable. So why is the downside so much less, what’s the message to investors on this dynamic?

Ernie Maddock

Yes. So I think it’s — what you’ve just described is descriptive of the general industry and maybe even a little bit more applicable to DRAM versus NAND which is a younger technology but certainly with respect to DRAM over the course of its history which is now some 30 years old as you’ve pointed out. You have seen a dramatic reduction in the length and the depth of many of the cycles. I think there are two things that’s attributable to.

First, we’ve seen a consolidation of the industry over that time if you back up 20 years, you may have had a couple of dozen industry participants. You had the Taiwanese enter the DRAM industry as late as the mid-2000s, and today there are three. There is Micron, and Samsung and Hynix. And as a result of that I think that you see more disciplined exercise in terms of how the industry looks at supply and I think more importantly on the demand side, everything that we’re seeing today in terms of trends relatively to artificial intelligence, self-driving cars, machine learning; you know, just tick of the list of things, we’re very topical today. Those are actually very significant drivers of both pieces of our business.

All of that data that’s being collected about you and everyone in this room is stored somewhere, that’s an SSD play in a data center and everything that is stored is eventually processed and tried to turn into information to offer someone competitive advantage and that’s a DRAM story. So I think you have a fairly robust demand environment, the CAGRs we’re projecting for DRAM for the next few years are somewhere in the 20% range for NAND, upwards a 45% and that robust demand environment in the face of fairly thoughtful investment is creating the opportunity for different cyclical behavior of the industry.

Unidentified Company Representative

So you have some new and exciting demand drivers at the same time that the consolidation is causing more discipline. And then on the supply side, you know, I thought it was very interesting on your Analyst Day presentation you show the drivers of — you know, you show capacity as a driver of bit supply growth, you show technology as a driver of bit supply growth for both the NAND and DRAM markets but for the first time that I have noticed, you’re showing throughput as a negative driver of capacity or bit supply growth. So can you help us understand what is that dynamic?

Ernie Maddock

Sure. Using just simple math for an example, if you’re producing 100 wafers at 20 nanometer DRAM technology node and you transition to a 1Y technology node, you may get 25% to 30% more bits on each wafer but as you might imagine, because that equipment is having to etch or deposit or etch more — finally, a texture geometries, the actual amount of time that a wafer may spend in an etch process or deposition process in aggregate would go up and as a result of that the same amount of equipment is able to etch fewer wafers albeit with more bits per wafer and so that throughput detriment that you saw on the charts at Analyst Day is representative of that and of course the way the industry copes with that is by adding the amount of wafers necessary to keep that wafer throughput flat.

So if as a result of that a 100 wafers turns into 95 at the next node, you have the option of just accepting that and you feel still have bit growth because each one of those wafers is producing more bits or you can buy the equipment necessary to keep output at a 100 wafers.

Unidentified Company Representative

I think there is a deep ceded suspicion as we talk about the competitive dynamic. So it’s a deep ceded suspicion in the minds of many investors that China will come in and then ruin — get into the market and ruin the pricing environment shortly after they get up and running just to gain share and dominate the market. And then there is another school-of-thought and the debate that says they just don’t have the right intellectual property to have a material impact on the market. So if you worry about the competitive dynamics in China and how should — how do you suggest investors think about that?

Ernie Maddock

So you’re paid to worry and if you do what I do for a living so I would never say that we’re unconcerned about that by any means but I think the point about IP is very legitimate and very powerful. If you look at every entrant into the memory business at some point in time, they had a legitimate resource core of IP that allowed them to form the capability around which they have advanced their technology and even with that as we just talked about, you had 35 industry participants in DRAM and today you have three. So even with — you know, legitimately acquired IP and assistance from the person who provided that IP or the entity that provided that IP you still had a larger number of failing companies than you had successful companies. If you add to that challenge the idea of no legitimate source of IP because the only IP holders in DRAM are the three industry participants and in various forms and flavors we have all indicated that we would be very challenged to think about transitioning that IP to someone who might think about industry capacity in that way or subject to regulatory approval, certainly as a U.S. based company, Micron would be subject to regulatory review of that.

It makes — what is a difficult challenge, extraordinarily difficult, so while we would never say that it would never happen, I would say that without a credible source of IP, the odds are very, very slim and even if you achieve a specific result; this is an industry that moves forward at [indiscernible]. So over the next three years, we will likely have two to three generations of NAND technology that will be introduced into the market by the industry whether it’s micron or any of the other participants. It’s a reasonable to think you will have two DRAM technology notes occur over the course of the next two to three years and that is a rate of speed that is very different than in any other industry that you can think off and one that has proven very challenging to successfully enter. So I think there are formidable challenges even with the appropriate support of IP without that it’s very, very, very challenging.

Unidentified Company Representative

And then on the other side of that coin, in China what is — what — do you have a particular — what’s the strategy for being highly relevant in that market, are you willing to gauge in JVs or license agreements or other partnerships?

Ernie Maddock

Well, that’s a wide range of outcome. So in terms of partnership, I think Micron has consistently been willing to partner with other companies, other industries, other governments where both parties derive benefit from that arrangement. So it’s not for lack of willingness of Micron to do that, I think the real challenge — because you threw out IP licensing in there; IP licensing is very different than just a JV where you both fund the fab and you both share the output. So in that range of continuum, I’d say anything that involves the transfer of IP would be at the difficult end, anything that involved a normal sort of financial benefit and exchange for both parties feels very doable. The question is whether that is the mutual interest of the parties, right.

Unidentified Company Representative

Fair enough. So I have one more industry question and then I want to drive down into the demand drivers. On the industry side, Toshiba’s NAND business is up for sale and I understand that there are lots of permutations of how that plays out. Is there a framework that you can share with us for how we should think about the opportunity or risk for Micron with Toshiba?

Ernie Maddock

Yes. So I’ll just comment from a general industry perspective because it’s really not appropriate for us to talk about specific circumstances but the most important question is whether any acquirer will increase the level of CapEx that is currently planned by that joint venture and that would lead to an expanded industry supply which would have a disruptive effect on the market as we have described today. So I mentioned earlier, 45% bit growth, that contemplates our view of what each of the industry participants will be investing to enhance their 3D NAND capability. So if an acquirer ultimately expanded that, we’d have to take that into account relative to the industry supply dynamics. Provided an acquirer does not do that, we are relatively — we have a relatively benign view about the impact to the industry.

Unidentified Company Representative

Fair enough. Okay, on the demand side for NAND, solid state drives are becoming more competitive on a cost basis such as hard disk drives. You — I think you guys are forecasting that solid state drives within a NAND business go from about 40% to 60%. Can you talk about the cost curves here and that move from 40% to 60% of your NAND business in solid state drives? What is it assuming on those costs curves?

Ernie Maddock

You know, we have provided some general views of cost curves, so if you think about some — just in back at the envelope metrics, you typically would see cost reduction that somewhere around half to maybe a little more than half of bit growth. So as we’re forecasting a CAGR for 45% bit growth over the next few years you’d think about a cost CAGR that would be somewhere in the mid-20% range plus or minus a little bit. I think that it is important to note that that is not the only basis of comparison, so that maybe relevant for a PC where you have a directly substitutable product for an SSD versus an HDD but if you look at for example, main data center storage today even at the pricing levels of SSDs today you have about a 30% TCO advantage for a solid state drive based storage environment versus an HDD because things like power consumption, reliability, maintenance and service ability are dramatically different for those two product.

So you will see that as you are now penetrating sort of the upper end of the market because those markets have a broader basis of cost comparative than simply the price of the device itself and as the industry comes down that cost curve over subsequent generations, I think you’re going to see broader and broader substitution and eventually we all believe that the industry will reach a price point that which for example, there would be no PC, desktop, laptop; you know, anything along that line that would have an HDD in it and then you begin thinking about things like near line storage and [indiscernible] storage. And although that’s a little further on the horizon, the magnificent thing about SSDs is there is a very broad, well established market and this is really about the industry hitting its technology points and its cost — corresponding cost points to consume a larger and larger part of that market with SSDs.

Unidentified Company Representative

On the DRAM side I think a lot of people even now still think about DRAM equaling PCs and since PCs are declining, the demand for DRAM can’t be that great. How you’re forecasting the DRAM market to I think increase to 20% to 25% range over the next few years annually, where are PCs in that stack? What is driving the growth?

Ernie Maddock

So PCs today and I would combine the spot market with PC contract market, it’s probably less than 20% of aggregate demand. As you pointed out unit growth might be slightly positive this year but you know, over a trajectory unit growth is going to be flattish, maybe declining slightly; content growth is growing at very, very modest rates but the really exciting pieces of the DRAM market today continue to be the mobile market and the data center market, it goes back to what I said earlier about everything that’s been collected about you is not only stored but it is processed. That is causing — for example, server units which were growing modestly, maybe 5%. Content on average is going up about 40% and importantly, the — that content is growing not only with core — what we would call main memory relative to DRAM, there is an increasing amount of that that is going to high performance DRAM that looks a lot like graphics. So very high performance, high throughput, high reliability, and that appears to be a core growth driver for DRAM, at least for the foreseeable future.

As we think about the mobile business, you know, again, very modest unit growth but content growth growing 30% or so this year and surprisingly, although we all tend to gravitate toward the flagship phones, the biggest growth in the industry is coming from sort of less developed countries moving from feature phones to their first smartphones, so it’s literally going from no mobile DRAM to maybe a gig of mobile DRAM and a little bit of NAND coming along with that. So that will likely fuel growth there for some time and then further on the horizon you see things like autonomous vehicles, edge computing has become a popular buzzword where actually a security camera instead of simply recording your image; it’s actually doing some processing because it’s recording your image and it’s also instantaneously comparing you to some set of universe of interested parties and as a result of that there is computing taking place there which means there is not only NAND going into that camera; there is a little bit of DRAM going into that camera environment.

So everything that we are hearing and seeing today with respect to the explosion of the collection and use of data is directly impactive in a very positive way to both memory markets.

Unidentified Company Representative

You know, if we — if you look at graphics penetration into the data center for deep learning, machine learning applications. You know, one of the biggest questions I get is what is the — how much more DRAM is there on a card — graphics card that goes into the data center versus a server blade — standard server blade? What’s the answer to that?

Ernie Maddock

You know, it varies but it’s a multiple two to — two to as much as four, maybe even a little bit more than that and it’s very situationally dependent so I don’t think that there is rule of thumb, it is very specific to the purpose of that implementation and that installation but what’s really important is that there is more of that on a relative basis going in and that memory provides an opportunity for Micron which has a leading position in the graphics business to further differentiate ourselves in the context of a data center environment.

Unidentified Company Representative

And is that a higher performance memory that would go on to that card?

Ernie Maddock

Absolutely, graphics memory is characterized by a higher performance throughput, higher reliability, so it’s the — you know, the most stringently performing among all of the variety of DRAM that we produce.

Unidentified Company Representative

I’m going to pause there to see if there is any questions from the floor. Right here in the front, can you wait for the microphone please. Right here.

Question-and-Answer Session

Q – Unidentified Analyst

The three main DRAM producers like you mentioned, they all have NAND businesses but the converse isn’t true, the NAND — some of the NAND players don’t have DRAM businesses. Is there a synergy legacy between the two kinds of memory in terms of how you see the competitive environment?

Ernie Maddock

Sure. So if you think about something like the Chinese smartphone market, the vast majority of product that shipped into theirs in the form of — somebody called an EMCP which has some NAND and some DRAM packed together with a firmware — with some firmware and a controller in front of it. If you think about the automotive business, there is an interplay between NAND and DRAM in those businesses as well and that will likely be a trend that increases on a going forward basis as opposed to a decline. So it’s more and more of this idea providing a memory solution to our customers. And so if you are only in one of those two technologies, you obviously — if you want to participate in those markets you’re going to have to acquire your DRAM from one of the current producers.

Unidentified Analyst

Okay. And as a kind of follow-up to that, do you see the capacity for those two [indiscernible] like — if the two memories are on different cycles, can you move capacity from one to the other relatively easily?

Ernie Maddock

No, it’s actually getting more challenging because most NAND architecture now is based on a 3D and DRAM continues to be planner. So it’s absolutely possible to do a conversion but once you do a conversion you need to stay in that converted state for some years to have the economic return on that decision.

Unidentified Analyst

Got it, thank you.

Unidentified Analyst

Can you talk about automotive because obviously you have to qualify as you stipulated on one of your conference calls, and so it’s quite hot, surely to dislodge once you qualified it.

Ernie Maddock

That is absolutely true. Automotive design cycles, it’s reasonable to think about those in terms of a decade. So you’re in the process today of qualifying for automobiles that might be 2019 or 2020 and once you are qualified you have to provide very high assurance of supply over the course of a 10-year period and of course automotive products like graphics are some of the most stringent, they have very high temperature requirements, high reliability requirements and you must be qualified throughout the breadth of the companies operation from supply chain, all the way through the sales — to the sales process to serve that industry because of the safety and regulatory requirements.

Unidentified Analyst

So just following up on that one. How many OEMs you have now in which you actually qualified? [Indiscernible] that they are qualified in eight out of ten and [indiscernible] is probably in all of them. How many are you qualified?

Ernie Maddock

We typically do not discuss specific customers with whom we are qualified, we have about 40% market share in the automotive memory business, our next largest competitor has 20%. So we have a very broad qualification across the breadth of the supply base.

Unidentified Analyst

Alright. Going back to the issue of capacity now, clearly the profit pool now is controlled by the Koreans on the DRAM side [indiscernible]. In the NAND Flash, the picture doesn’t look that great, you know. Assuming that there is an [indiscernible] from Toshiba, do you think that the industry could return to some sort of order during the next two/three years, you know?

Ernie Maddock

Well, it’s not my belief that the industry is in disorder now. No, I was specifically talking about NAND; so you have four sources of IP with six industry participants. If you actually look at the profit pool, it has expanded just as rapidly for NAND as it has for DRAM and the Toshiba situation really has not caused a disruption in that profit pool. So I think the most important thing to remember is as a technology, NAND is about 15 years old and DRAM is about 30 years old and it is going to take some maturing of the NAND market which will occur over the passage of time to have a clear review of what the profit pull potential could be but there is no reason to think that it wouldn’t be equivalent at some point to DRAM. You’re starting from a much better position, right, when DRAM was as old as NAND, there were double the number of industry participant. So we actually are very constructive on our view of what the profit opportunities will be with NAND.

Unidentified Analyst

Thank you.

Unidentified Analyst

I think this will have to be the last question.

Unidentified Analyst

[Indiscernible] space, I mean what’s the typical DRAM and NAND content per car running at? How much is it growing — we know about PCs and phones and we’re used to those but cars not so much.

Ernie Maddock

So we think that the automotive content probably is going to grow over the next three years, 300% to 400% today, you might see two to four gig of DRAM that’s going to be some more closer to 16 and up depending on the car. If you look at NAND today, you have 16 to 64 gigabytes of NAND, in the future you’re going to have maybe a terabyte of NAND if you think about a fully autonomous vehicle, right, it has to store all of these maps and process set in real-time; so it’s a very, very significant content growth. And I think the other thing to bear in mind and it’s similar to the comment I made about growth in mobile NAND; you will definitely get an uplift from the very high end fully autonomous vehicles but don’t forget that there is a very large opportunity with cars that today have zero to very little DRAM content that in the future there will be DRAM and NAND content in every single vehicle whether it be to power a backup camera or whatever that maybe. So the opportunity is not solely limited to the very, very high end fully autonomous vehicles.

Unidentified Company Representative

And I think that’s going to have to be the last word. Ernie, thank you very much for joining us.

Ernie Maddock

My pleasure.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY’S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY’S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY’S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!



Source link

Recommended For You..