Seeking Alpha
View as an RSS Feed

H. Bruce Campbell  

View H. Bruce Campbell's Comments BY TICKER:
Latest comments  |  Highest rated
  • Intel: The Memory Company [View article]
    Some valid point taken StrategicInvestor. But I disagree with several key elements.

    Regarding Russ's projections, I don't perceive him as attempting to paint a portrait of a potential future in colors of certainty. In large measure he's simply sketching possible financial fundamentals in beloved wafer math which many of us then weave into our judgments about possible future developments, both technical and financial. Russ is perfectly aware of the massive number of moving parts which exist in the processor / memory / storage technology industry, and has never suggested that any projections are more than forward looking speculation, as he's often acknowledged.

    But I agree that there's a popular tendency for many investors to try to characterize future developments in terms of past corporate style and focus, a myopic error in judgment in my view. For my part I differ with both of you regarding how Intel (and Micron in large measure) will pursue 3D NAND and 3D XPoint.

    In looking forward we must remember that IMFT's version of 3D NAND is a unique technology. And nobody appears to be even in the ballpark of developing a peer to 3D XPoint. And IMFT might possess very tough IP moats for these technologies - perhaps particularly in the case of 3D XPoint. Given that:

    While I suspect Intel will, as usual, extract as much blood from early adapters as possible, in a break from tradition I doubt that'll be a serious focus - my sense is that Intel intends to out Korea / China the Koreans and Chinese with 3D NAND and 3D XPoint by setting their primary focus, by far, on rapid market dominance. My guess is that they have no intention of lingering with either of these products by setting limited production volumes or pursuing further development at a leisurely pace, but rather they intend to chase higher nodes (in both layers and shrinks) very aggressively - Samsung style aggressively - and equip all the fab space at their disposal, and then some, in a sprint to very high level manufacturing capacity.

    Technology coups are fleeting. Intel's keenly aware of this - the firm still suffers from portables market burn scars, and has no intention of allowing the mass market for 3D NAND or 3D XPoint to slip through their fingers due to a flawed notion that they should focus primarily on the high margin market for an extended period, or simply due to execution pace laziness. My sense is that Intel's in a tightly focused mood to flood the world with 3D NAND or 3D XPoint, and will do their best to make that happen.

    Russ observed "While Samsung's output is mostly memory, Intel cannot fall further behind on total wafer output or they will eventually lose their sales leadership and maybe even their manufacturing technology leadership." I heartily agree. Which leads me to disagree with his timing when he later states "So, 3D NAND should lead to new Intel business of $45 billion to begin with the expensive "low-hanging-fruit" of the data center in 2016 and progressing to the client business by 2018."

    My guess is that Intel has no intention of lingering to 2018 before aggressively addressing consumer level applications for 3D NAND and 3D XPoint. I suspect they'll charge after that sector with fire in their eyes and intent to flood the planet with 3D NAND and 3D XPoint as quickly and thoroughly as possible. They don't underestimate Samsung - they know that if they linger or fail to execute at a wartime class pace, Samsung will find some way to steal the thunder from 3D NAND and 3D XPoint. And they have no intention of allowing that to happen, necessary sacrifice of significant margins notwithstanding. Intel will, in my speculation, pursue this as an immensely high volume business. If well executed, they'll still earn a ton of money in spite of more modest margins. And they'll dominate the storage market, which is the biggest prize, especially considering that intimately connected storage and memory are quickly becoming the most critical ingredient in best of class IT technology, whether high end or consumer sector.

    My sense is that 3D NAND will rather quickly fade in this progression as well - I suspect storage will become, overwhelmingly, a 3D XPoint story. NAND, whether 3D or otherwise, will fall into obsolescence in large measure, doomed by inherent performance limitations just as is occurring with hard drives. And in the case of 3D NAND, its death might come more quickly. Hard drives still offer one advantage: Dirt cheap cost per bit. Yet they'll still mostly perish. And it's not clear to me that 3D NAND will, within one or two years, offer any advantage at all, even including cost per bit, over 3D XPoint, in part due to volume related efficiencies which could offset additional fabrication steps (if any and if significant). And 3D NAND substantially under-performs 3D XPoint in key areas. So in my estimation it's doomed.

    Intel isn't what is was, nor will it be what it is now. Nor will it become like anything in the ARM ecosystem. Intel's changing into something its never been in a very highly focused effort to fully leverage 3D XPont (and, briefly, 3D NAND) so as to outflank Samsung to the vast bulk of the storage market (wherever it resides, from within an SoC all the way to attached storage). And within Intel this vision is very clear and tightly focused.

    And in effect Micron will follow suit as a constructive partner, and also profit richly.

    All just in my personal speculation of course - nothing more.

    Highly concentrated long Micron common and calls, and Intel calls.
    Aug 26, 2015. 02:14 PM | 16 Likes Like |Link to Comment
  • Reports: Intel will have to wait until 2016 to take iPhone share from Qualcomm [View news story]
    And of course "...handing out jobs to women and people that aren't qualified." is broken grammar too. ('...handing out jobs to women and people who aren't qualified.' is of course the grammatically correct, though still sexist and evidently baselessly discriminatory, phrase.)

    If TheHash has good reason to believe Intel's performance is suffering due to unwise priorities in their hiring practices, he should make a credible case. My mind is open to well reasoned and analytically supported arguments. But baseless empty rants are of no use to anyone here - they're just rhetorical clutter.
    Aug 25, 2015. 03:07 AM | 5 Likes Like |Link to Comment
  • Intel And Micron: More On Memory [View article]
    Always a treat Mr. Fischer! I love the "Intel is and always has been a memory company." theme - it's so clear, but I didn't see it clearly until you blew the fog away to reveal the simple yet previously obscure reality. Intel is indeed a memory (and ultimately storage) company.

    Considering leveraged technology in total, my guess is that 3D XPoint will, for Intel at least, be proportionally akin to the iPhone for Apple. Possibly even more so if IMFT possesses a really tough IP moat around the technology, an advantage Apple didn't substantially possess. Most people didn't anticipate the massive scale of the iPhone's success within the first year or two after introduction. But Steve Jobs was confident and made the right investments. And my sense is that 3D XPoint might follow a similar public level success recognition path. But public reservations notwithstanding, I suspect IMFT brass are confident, and will pour CapEx into it more aggressively than most of us expect.

    Focusing on just the storage side of this story for this comment:

    I just purchased my first real SSD yesterday. Generally I love new toys, but in this case I felt no sense of excitement - instead it was a melancholy chore. I don't want NAND storage. Why should I? By credible accounts 3D XPoint is immensely superior. With all due respect to the charm of dirt cheap prices, when a performance chasm as mammoth as seems to exist between NAND (including even disruptive cost IMFT 3D NAND) and 3D XPoint, I'd never purchase a NAND based SSD if a 3D XPoint based SSD option were available. A 2.35 price multiple wouldn't deter me - the performance difference is far greater, so to me a double or even triple price is tolerable. Particularly since the long term reliability of my precious data is a huge factor. Besides, the operational savings, including total direct and indirect power and heat related ancillary costs (such a system longevity), would offset some of the initial cost difference over the life of the storage device anyway.

    I don't want NAND storage - I want 3D XPoint storage. And I suspect some in these halls already feel the same way. And far more importantly, after the full mass of the global geek culture's opinion weighs in with complete visibility, within perhaps a year or two, very nearly everyone will shun NAND based storage - they just won't respect it when such a superior option becomes available. Who wants a flip phone when iPhones are available? Screw the price difference when performance chasms are that wide...

    My guess is that IMFT anticipates these factors already. A hint seems to me to be in the original 3D XPoint presentation. The following question and Rob Crooke's answer occurred at about 36:08: [Question:] "...do you see this having a big affect on solid state drives as well?" [Answer:] "Uh, it'll be, uh, you know really fast, uh, storage uh capability, and so it could attach, either from a storage or memory attach point, so for the highest performance storage it'll be a great alternative."

    And that ignores the also chasm class endurance advantage. And others. Flip phone or iPhone - which do you want, price difference notwithstanding?

    I don't mean to suggest 3D NAND will wither to zero short term. But I do perceive a disconnect between Rob Crooke's comment above and their clear statement in the same presentation that 3D XPoint won't affect their 3D NAND road map. And I suspect IMFT brass are well aware they're presenting at least partially conflicting projections for 3D XPoint and 3D NAND. The presentation was very hastily arranged, a lingering mystery. Perhaps IMFT didn't really want to announce it, but had to for some reason. But they still want to try to shield sensitive logistics. Samsung won't just sit on its fanny while being run down of course - they'll counter as best they can. So big wide IP moat or not, if 3D XPoint is very close to ready for prime time, IMFT will, if wise, saturate the market with it as rapidly as possible.

    Sure they'll extract huge margins from early adopters, and only then move to moderate margin markets. But I suspect it'll be an unusually fast process - I doubt IMFT will linger on fat margin markets while Samsung works 24 hours and 400 mG of caffeine per day in a war climate frenzy to find an effective counter. With a $65 B market dangling, IMFT won't, if they're wise, give white hot fire in their belly Samsung any extra breathing room - I suspect IMFT, leveraging their Western Digital and Seagate partnerships, want every penny of that $65 B in their coffers, and not a single Won in Samsung's coffers. Samsung's already plenty tough to keep pace with sans a bunch of extra future R&D money in Seoul after all...

    At the moment I'm most concentrated in Micron by a large margin. I'm unlikely to tilt more toward Intel near term even though I currently take no issue with Russ's sense that Intel will be the bigger 3D XPoint winner. In part due to Micron's Angstrom sized P/E, but also because I suspect Micron will rise before 3D NAND, 3D XPoint, or their DRAM node improvements make a significant financial impact - the market won't remain blind to what will unfold forever - in time anticipation investments will begin to trickle in, then establish momentum. And I'm fond of the notion of new excitement finding focus on a firm with such a minuscule P/E.

    And not so late as we're being led to believe in my estimation - I suspect Micron and especially Intel intend to be much more aggressive about saturating the storage market than they're revealing, and that CapEx evidence of their intentions will subsequently trigger the beginning of momentum investing. Sooner than many expect...

    And to reiterate from my earlier comment, Apple strikes me as a potential 3D XPoint early adapter wild card. (Perhaps paying rather dearly in a cute role reversal.) If so the moment the market whiffs a hint of it we'll see the obvious stock price reaction of course.

    All just my personal novice speculations of course...

    Highly concentrated long Micron common and calls, and Intel calls.
    Aug 18, 2015. 05:28 PM | 4 Likes Like |Link to Comment
  • Intel, Micron: My Thoughts On 3D X Point Memory [View article]
    Thanks Russ - straightforward and compelling as usual.

    But I suspect more layers are at work in the IMFT relationship, in part because I doubt either partner would have supported the long research effort in earnest had it been clear that the fruits of the work would ultimately harm them materially. So I suspect Intel and Micron agreed to something which keeps the relationship healthy and each firm whole.

    In Stephen Breezy's latest article at http://bit.ly/1M0XUrV, William Tidwell commented:

    "ARM licenses are doing pretty well these days - surely there would be a market for 3D XP there? Is it possible that Micron becomes an ARM licensee in order to sell an integrated on chip package?"

    I don't mean to suggest that I think that possibility might be part of a plan to keep IMFT healthy, but it's an interesting point which at least illustrates that 3D XPoint might spawn any number of heretofore unlikely pursuits.

    Most others seem to disagree, but I'm juiced by the possibility that 3D XPoint will ultimately replace 3D NAND, irrespective of IMFT's insistence that their 3D NAND road map remains fully intact. As you said "Two shrinks and 3D X point begins to crowd even 3D NAND." And while it's said to cost more than 3D NAND, there are a world of questions about that metric in my mind - is that fabrication cost or planned initial retail price? And if fabrication cost, is that now or after a couple of shrinks? And is it padded more that whatever padding might be in place for 3D NAND?

    To me fabrication cost would have to remain significantly more expensive to offset 3D XPoint's clearly superior performance in speed and endurance over 3D NAND. And though I don't have a tangible basis to judge, I doubt a sufficiently substantial fabrication cost difference exists. If any at all, particularly after a shrink or two.

    My sense is that 3D XPoint's early limited output will be dedicated to displacing DRAM as most describe since in that role Intel can reap superb margins on uniquely high performance SoC / memory / SCM modules as you describe. But as output ramps up (particularly after a shrink or two) both Intel and Micron could pursue the very roughly $65 storage market aggressively, sacrificing 3D NAND in the process.

    Based upon the metrics I've reviewed, I'd much rather have 3D XPoint based storage in all my personal cyber systems. If others feel differently, I'd sure like to hear why. If not, for about half a $65 B pie, maybe Micron will be perfectly happy with 3D XPoint irrespective of the fate of DRAM...

    Highly concentrated long Micron common and calls and Intel calls.
    Aug 3, 2015. 12:55 PM | 8 Likes Like |Link to Comment
  • Micron: Not Selling Hot Dogs Anymore [View article]
    To correct an egregious oversight in my first comment here: Stunningly informative article Stephen - a dazzling display of refined exploratory insight! (Purple Swan too of course.) Thanks so much! And thanks Jordan and HenleyTheCat - very kind!

    And Cincinnatus, my sense is that there's very little space between our views. My original intent was to generally avoid certain specifics in the context of an effort to support an enthusiastic heart during a time of personal and professional adjustment. I agreed with your points, and for the most part do in your response to my first interminable comment. I suspect they're well understood and respected by Bruce B. as well, and that he's adjusting his views and style in response. I simply hope the process of adjustment is painless in part because his core enthusiasm is important to many of us here - he's a good information digger (far better than me, as are so many here).

    He's not alone in occasionally overreaching or responding to frustration by attaching blame when in fact no sin was committed, but I suspect self examination is occurring as we speak, and in the meantime my sense is that most here are independent minded and think for themselves, and thus filter out most overreach and other somewhat minor sins rather easily. Whatever the case, personally I still need the diggers here - they provide the core grist required to formulate better investing thesis.

    "Neither of those statements are realistic." Agreed. No corporation is our investing mother - generally they'll be transparent when they reasonably can, but coy when any of a variety of conditions demand it. Investors must accept that rather simple reality, then try to discover as many hidden truths as practical. In my view Intel and Micron are committing no sins in the context we're discussing.

    And I respect the frustration behind: "... people on SA seem to be claiming much more of Xpoint than either Intel or Micron did." And yes, whether on this cycle or next, the misplaced anger from less seasoned investors you characterized in your last paragraph will occur, sorely exacerbated by loose use of superlatives by those of us who should know better. I dislike having to endure such end point reactions too.

    But there's a flip side to that quicksand in our park - sometimes the opposite mistake is made - sometimes, in the spirit of disciplined conservative realism (whatever that means), we fail to adequately recognize unique opportunities and reasonably alert our peers. Agreed that's not Micron nor Intel's job. But when we believe in our hearts that change might be swift, we in these halls ultimately have a duty to convey such impressions transparently. Sans superlatives and with disclaimers of course. But still transparently.

    My personal sense thus far is that "Nor should one expect products sooner rather than later out of XPoint." might prove to be too conservative. Again agreed that Intel and Micron aren't claiming nearly as much from 3D XPoint as some of us. But I don't see any role transgressions there. Intel and Micron dare not state, at least openly, that superb success is coming. But, sans superlatives and with humility, those of us who believe in our hearts that the evidence suggests it might be so must convey such thoughts openly so they can be effectively debated.

    In particular "Intel and Micron are only hinting at products in 2016, and they're being cagey about it..." seems true and is important, but I interpret the matter differently. I suspect Intel, which isn't the least bit shy about extracting huge margins from customers who desperately need a serious performance edge only Intel can provide, will do just that with early 3D XPoint enhanced Data Center Group targeted systems (and other specialized performance critical niches). Early customers will pay very dearly, and Intel won't address wider markets with less extravagant margins with their limited quantity early production. Thus (in part) Intel's being coy. But that doesn't necessarily mean their bottom line won't be juiced.

    My sense is that there was a lag time in our collective recognition of Intel's Data Center Group's level of contribution to the firm's financial performance. And we still have a tendency to focus too much on the massive mobile market during times of technology transition.

    I suspect Intel and Micron will ramp 3D XPoint as quickly as they can, and that we'll see financial impact sooner than their public posturing suggests. But perhaps not in mobile near term. But when analysts perceive the juiced financial results in the Data Center Group, then extrapolate them to mobile, if Intel and Micron still have a substantial IP moat those analysts might begin to transmit more excitement and raise price targets, leading to the obvious shorter term effects on their stock prices and P/Es.

    Just my personal speculations of course. And a thousand other things could happen, some good and some not. Apple strikes me as a potential 3D XPoint buying wild card for example (but if so they might pay rather dearly too in an interesting role reversal).

    So perhaps we differ in our moderate term expectations in some measure. But otherwise my sense is that there's not much space between our thoughts.

    Cheers with respect!
    Aug 3, 2015. 02:39 AM | 5 Likes Like |Link to Comment
  • Micron: Not Selling Hot Dogs Anymore [View article]
    BioLookX: Do we know that they're not?
    Aug 2, 2015. 04:35 PM | 1 Like Like |Link to Comment
  • Micron: Not Selling Hot Dogs Anymore [View article]
    Valid points Cincinnatus. But on the whole Bruce Burnworth's investing logistics, as best I've perceived them, haven't been stubborn nor ill informed - he has critiqued himself transparently in these halls, and like many of us has erred on both the aggressive side and the conservative side. As best I recall he openly revealed selling Micron calls much too early after the Elpida acquisition for example.

    Those of us who embrace more of a swing for the fence philosophy know of course that we'll often miss-time matters or be outmaneuvered by difficult to predict events, sending our precious call money down the swirl - we all have numerous battle scars. But we soldier on, struggling mightily to try to find optimum levels of aggression and timing whenever significant new developments become visible.

    And this is certainly an interesting and challenging point in time. We can never know just when the pendulum has reached its zenith, but it can be argued that one clue is an investing community which begins to express skepticism that it'll ever reverse, or do so only slowly.

    And Bruce clarified that some important features in Windows 10 won't be available in aging IT hardware, which invites some interesting speculation. Buzz Aldrin wisely lamented that he was promised Mars, but instead got FaceBook. In my view FaceBook is about the most insipid human folly to ever spring from the most shallow enclaves of the human heart. But sadly we're awash in it, and it's easy to imagine that just one significant gesture control or similar 'enhancement' of FaceBook will infect many with 'New PC' fever, pouring enough of their money (or more likely their debt) into IT store cash registers around the globe to move earnings needles. (Mars can wait!) Such are our times, alas...

    Pendulums do reverse course, and sometimes substantial lubrication is added to their bearings by new technologies and fads. Bruce can ably speak for himself, but my sense is that his sense is that current conditions might be signaling the beginning of a significant pendulum reversal. Likely not in time to rescue a bunch of January 2016 calls, including some of mine. So be it - we might root with open enthusiasm for a last moment miracle, but we always knew the risks perfectly well, and will shoulder our losses with grace just as we celebrate our wins with restraint and respect. Having said that, while the market is often terribly myopic, on rare occasions it experiences a genuine forward looking epiphany. And anticipating the anticipation of other investors is an important facet of call option logistics, so we who use them to swing for the fence can be forgiven for trying to peer a bit further ahead and openly consider and debate both best case and worst case scenarios in an effort to perceive market epiphanies before they occur.

    And personally I'm developing a sense of time warp. Up to almost now, this year has been awful for Micron investors, and mediocre at best for Intel investors. Not just due to floundering profits, but also due to an appearance, in some measure, of major technology development malaise. But obviously important work has been occurring under investor's radar. And perhaps more than we realize thus far - perhaps Intel has already integrated 3D XPoint into major products in development labs, and perhaps some will be introduced surprisingly soon. I'm just speculating of course, but the relatively slow perceived pace of the last six months seems to me to have distorted our sense of the real pace of underlying progress. If so, the pace going forward may be perceived as unusually fast by the investing community. Which might juice the anticipation factor in the stock prices.

    January 2018 derivatives will become available relatively soon, and I'll shop very eagerly for perceived bargains among them. My sense is that there will be substantial developments in IT, robotics, AI, and life sciences which, in some cases lubricated by 3D XPoint, might be of sufficient magnitude soon enough to make 2018 calls seem especially alluring. I'm also inclined to think many 2017 calls will prove to be lucrative - I suspect the lull in consumer IT sales will reverse in some measure relatively soon. And even if sales simply revert to normal levels, Intel and Micron are currently priced roughly as if depressed levels will last indefinitely. So it's only natural for an aggressive investor to smell opportunity here, particularly when stimulating new technology's being added to the equation.

    Agreed that superlatives are frequently misused and can frustrate even experienced investors. I'm guilty too, but trying to reform myself. But there is an element of existential joy in investing endeavors too, and I'd hate to see these halls become too conservative, gray, and mundane. If so we'd lose a lot of the joy of investing, and also a lot of important exploratory information content. So while I agree that superlatives are often improperly employed, I never felt an urge to kiss a date with a mundane banker personality irrespective of her beauty, health, good hygiene, youth, etc. But add Bruce's intellect (in female form factor please), even if it's a bit over the top at times, and enthusiasm, and I'm all in.

    There is a kind of romance within the spiritual element of investing, and wisely leveraged it can be immensely constructive. And it's quite alive and healthy in these halls. I hope we can keep it that way.

    A separate thought spawned by the debate with g3: I have to wonder what level of performance IBM's Watson will achieve once optimally enhanced with 3D XPoint technology. Particularly if Watson's redirected to a securities management related role. HFT and bots certainly aren't the end story to cyber enhanced investing - greed being universal, far more serious AI will be employed, and perhaps rather soon. Change often seems slow, but sometimes it happens very fast. Especially if lubricated by either greed or war...

    Highly concentrated long Micron common and calls and Intel calls.
    Aug 2, 2015. 04:16 PM | 20 Likes Like |Link to Comment
  • XPoint From Micron And Intel: A Strategic Coup [View article]
    Thanks again William regarding "I'm doubtful about the 3D NAND impact."

    I lean the other way, but believe it's far too early to be substantially confident. My sense of Intel's view comes primarily from Rob's comment at 36:08 in the presentation. And though development of 3D XPoint was immensely expensive, I heard nothing to the effect that ordinary fabrication of 3D XPoint will be expensive. (And it's not obvious from the structure that it should be, though I grant it's impossible to draw such a conclusion from a simple presentation slide.)

    My instinct is that Intel especially feels they have a solid IP moat with 3D XPoint, and, considering its superb storage performance, will fully milk it with very lofty margins, tapering them only in future years when other competitive pressures begin to build. If that's about right, it implies that 3D NAND will be the technology of choice for ordinary applications simply because quite high retail prices for 3D XPoint storage will discourage its use by any but the richest hands with the most critical performance needs.

    Having said that, I don't know the basis of thinking by respected analysts - they might very well be privy to facts I've missed. But the 3D NAND story's been out much longer and is pretty solidly understood of course. And thus a bit entrenched - the story carries a bit of momentum. So if the raw reality is that 3D XPoint is an even better storage solution overall, it might simply take a week or two for analysts to reconsider 3D NAND's true position alongside ever so fresh 3D XPoint.

    Just my instinct at this stage...
    Jul 30, 2015. 03:35 AM | Likes Like |Link to Comment
  • XPoint From Micron And Intel: A Strategic Coup [View article]
    Thanks William! It's hard to imagine a more appealing embellishment to the HMC (Hybrid Memory Cube) - inclusion of layers of 3D XPoint seems like a perfectly natural fit for HMC to me too.

    And yet, a fan of TSV (Through Silicon Via) and HMC notwithstanding, my current view is that Intel has other appealing ways to achieve a highly intimate SoC to memory and storage connection as well, including optical (if it's sufficiently developed).

    Currently I don't feel well informed enough to guess which connection method Intel thinks will perform best for any particular class of SoC though. (Though it's tempting to think there'll ultimately come a day when even substantially different fabrication processes will be layered in direct silicon on silicon 3D structures - someday...)

    But at this time perhaps Micron's options for accomplishing highly intimate connections with SoCs fabricated by firms other than Intel begin and end with HMC - if they have other options, I'm not privy to them. Thus perhaps 3D XPoint will amplify HMC's role significantly. Perhaps...
    Jul 30, 2015. 02:40 AM | 2 Likes Like |Link to Comment
  • XPoint From Micron And Intel: A Strategic Coup [View article]
    If you're interested in the IP question, I recommend going to precisely 31:00 in the webcast and watch Rob Crooke as he says:

    "Yea I agree and it's - you know as far as being in the memory business, ah, it's good to have something other people don't have, that's for sure."

    Look at the wide grin on Rob's face at the end of that comment - it's classic. I'd swear he's just one step away from looking into the camera then saying, 'Hey competitors...' while putting the tip of his thumb against the tip of his nose and waving his fingers back and forth in the air. It's a truly great grin...

    Combined with the later comment I mentioned above, my sense is that they feel they have a wide and solid moat - I think they're confident that 3D XPoint or similar won't be emanating from anyone but Micron and Intel for a long time. There's no definitive indication I'm aware of, but the subtleties of the webcast suggest to me that it's so.

    They also said (I don't recall when) that 3D XPoint won't affect the 3D NAND road map - they were very clear about that. My guess is that it could on a technical and maybe even cost basis - it seems clearly superior in every performance category, and though I have no sufficient basis to judge yet, my guess is that it's not intrinsically more expensive than 3D NAND to fabricate, or at least not substantially so. And recall Rob's comment at 36:08 - he all but said 3D XPoint will be the mass storage tech of choice for applications which require peak performance.

    So my guess is that Micron and Intel feel nobody else can touch this technology, so they're at liberty to leverage thrilling margins out of it until all the rich hands have ponied up a lot of dough to acquire it. Then they'll leverage all they can from the next tier of wealth and so on until, finally, the time to start displacing 3D NAND as a matter of consolidating products for industrial efficiency arises. And they'll still make a bunch of money so long as their moat remains intact.

    It's a very rosy picture for Micron and Intel, suggesting reality can't possibly be so pleasing. And yet, for me at least, multiple reviews of the webcast seem to reveal enough connecting hints of a solid IP position for a very solid technology coup. It's hard to imagine a more appealing situation (except to have clear proof of each factor).

    It should be fascinating to watch Intel's stock performance today. I don't see much reason for skepticism in this development, and I suspect multiple analysts will start exuding excitement quite soon. In my view Intel should have risen at least 10% yesterday. But I suspect high excitement lag time won't last long...
    Jul 29, 2015. 07:35 AM | 8 Likes Like |Link to Comment
  • XPoint From Micron And Intel: A Strategic Coup [View article]
    Thanks Jordan, very kind.

    On the IP front, I listened to all but the last 7.5 minutes of the webcast again. At about 36:30 a summary of IP questions was asked, with Rob Crooke of Intel responding "Yea uh, today we're really talking about the ... the technology itself and what its capabilities are, and this was a jointly, uh, developed technology that we both, uh, have rights to and beyond that it probably doesn't make sense... [to discuss IP issues any further in this presentation]"

    Just speculating from his demeanor and wording, I suspect Intel and Micron believe they possess significant IP barriers which they aren't obliged to share with outside firms (including SanDisk / Toshiba). But I'm just purely speculating of course...

    As interesting was the question and answer (again from Rob Crooke) just previous, at about 36:08: "...do you see this having a big affect on solid state drives as well?" "Uh, it'll be, uh, you know really fast, uh, storage uh capability, and so it could attach, either from a storage or memory attach point, so for the highest performance storage it'll be a great alternative."

    That seems to imply displacement of 3D NAND storage, initially in higher performance applications. And given both firm's excitement about the scalability of 3D XPoint in both x/y and z, plus the superior endurance claims, it seems to me that it might displace all NAND in time. Rob Crooke seemed to me to leave that interpretation open...

    So if 3D XPoint proves to be as solid as hoped, it seems to me that it'll simply render NAND, including 3D NAND, ultimately obsolete. (Assuming 3D XPoint can scale vertically as economically as 3D NAND. But at the moment I see no obvious reason it shouldn't).

    If 3D XPoint proves to be as solid as hoped, the TAM (Total Addressable Market) would seem to be, figuratively speaking, almost open ended. So I can't help but wonder what kinds of frantic CapEx and facilities scrambles are occurring at Micron and Intel these days...

    And to indulge in a really big leap of speculation, I wonder what sorts of corporate maneuvers will ultimately occur to put facilities owned by other firms into productive fabrication of new rather than suddenly obsolete technology...
    Jul 29, 2015. 05:16 AM | 10 Likes Like |Link to Comment
  • XPoint From Micron And Intel: A Strategic Coup [View article]
    Wow, great review William! Hastily composed, yet smooth as silk as if labored over for an extended period. Thanks so much!

    I'm not privy to the facts, and hope RSA or others will weigh in, but I'm suspicious that IP cross license agreements among these parties are limited in scope to preconceived technologies (though inclusive of modest evolutionary enhancements of them) - I suspect they don't cover substantially new technologies.

    I thought I heard a positive comment in the presentation about an IP advantage as well, though I need to listen to it again - maybe I misremember. (And I can't be definitive about any aspect of this matter.)

    Independently: I'm suspicious Intel has been developing means to include 3D XPoint into their current SoCs since last year. As bright minds here have noted, code to support NV memory in major OSs was released last year. In my estimation Intel can't have been sitting on their hands on the hardware side of the matter - I suspect variants of all their current SoCs which incorporate highly intimate XPoint, whether directly in the die, or so intimately attached as to have a very nearly full speed channel with XPoint, are being rushed to readiness. So I suspect more surprises are in the near term event queues. And that Intel's fortunes in this drama are being sorely underestimated by the investment community.

    Intel should have risen at least 10% yesterday too. In my insipid opinion... (But granted Micron's microscopic P/E is no small lubricant for a fast investment community reaction. With a 6 P/E on their pedestrian business, measured in a rather deep dip in a demand cycle, it's a whale of a challenge to find any credible down side investment risk to Micron at this time.)

    Highly concentrated long Micron common and calls, and Intel calls.
    Jul 29, 2015. 12:38 AM | 7 Likes Like |Link to Comment
  • Intel/Micron: 3D NAND And SSDs About To Explode [View article]
    One week ago the 21 August 2015 $19 MU call had 100,450 contracts, a $1.01 bid / $1.07 ask / $1.05 last. So in just that particular call there was a collective bet of over $10 M that Micron's stock price would eclipse about $20 rather quickly.

    It's now $1.43 / $1.51 / $1.44 (which seems fairly cheap to me under the circumstances), and up to 101,030 contracts.

    A whale of a lot of money's been riding that single particular call option...
    Jul 28, 2015. 05:23 PM | 6 Likes Like |Link to Comment
  • Intel/Micron: 3D NAND And SSDs About To Explode [View article]
    Odd that Intel's not responding more positively. If they could fabricate X-Point within an SoC (possibly with a 3D layer approach in part to address differing fabrication recipes), or use their optical interface technology to chat with X-Point at full X-Point speed, who could possibly compete with their SoCs in almost any application?

    Also interesting that the investment community couldn't see this coming. Both firms managed to clothe it, in some measure, as just one of several promising but not ready for prime time technologies for a considerable time.

    And also odd that the announcement was so hastily organized. Which seems to suggest that more might be afoot...
    Jul 28, 2015. 02:23 PM | 7 Likes Like |Link to Comment
  • Intel/Micron: 3D NAND And SSDs About To Explode [View article]
    I admire Samsung's belly fire and technology savvy. But all entities have some weaknesses which spring from their core characters. Samsung moves ahead with a wartime pace mentality - their sense of intense urgency propels them forward with remarkable speed and confidence. So they're a formidable competitive force.

    However... We who are fossil class engineers have all experienced the pain of confidently following a road with potholes we felt certain could be smoothed out nicely only to discover that they persisted no matter how determined our work to vanquish them. Every effort to patch seemed immediately swallowed as if the potholes were bottomless pits. Ultimately we realized we made a fundamental science or technology path selection mistake which couldn't be overcome by any known or contrived technology or was inherent in Natural Law related limitations. In such cases we finally had to face reality and abandon most of our invested time and resources - we had to start over with a clean sheet of paper.

    And those nightmares happened most frequently when we bolted into design and development projects with, you guessed it, unusually frantic pace.

    Nobody can know with certainty how charge trap technology will ultimately play out - maybe it'll be a road paved with interminable frustrations and irresolvable performance limitations which finally renders it a dead end. Or maybe some magic will be found which allows it to reasonably keep pace with or even outperform floating gate technology.

    But as an old engineer peering at the current evidence I suspect the former - I suspect those pursuing charge trap implementations of 3D NAND (or V-NAND) have a serious competitive problem on their hands which can't be resolved within the bounds of moderate or even rather long term technology, and probably mostly recognize it. (Even though they're likely making every effort to find effective means to overcome their potholes and thus vanquish a nightmare.)

    And IMFT seems to believe this is the current state of affairs. They revealed rather solid opinions to that effect as Phred, Russ, and others here have highlighted. They seem confident that their initial floating gate and relaxed node choice for 3D NAND was the correct one and that it's given them a substantial lead in a very roughly $65 B class unfurling market. They're transmitting their confidence rather clearly in spite of knowing the risks of visible overconfidence well. In my estimation they wouldn't display such open confidence unless it was quite strong.

    If they're right, Samsung and SanDisk can't simply jump technology ships and hope to catch IMFT anytime soon. Multiple very challenging process technologies must be painstakingly mastered for any firm to implement IMFT class floating gate 3D NAND - it's nothing like easy, and even wartime class pace won't enable Samsung to erase the time they lost by choosing the wrong road in the beginning (nor wartime class espionage in Hynix's case). And SanDisk / Toshiba seems to be in the same fundamental pickle. And there are IP barriers too of course.

    Granted anything can happen - no future is assured. But we who swing for investing fences don't seek assurance at times of investing opportunity, we simply seek compellingly attractive odds. Waiting for proof imposes very high odds of missing the bulk of possible investment rewards.

    In this arena the investing challenge is more taxing for those with limited technology backgrounds. But the window is nonetheless open for anyone with a free and exploratory heart and a willingness to listen carefully to the immensely generous and very well studied words of Stephen Breezy, Russ Fischer, Electric Phred, William Tidwell, and many superb others. And those who offer contrary opinions too of course.

    In this case the picture seems unusually clear to me. IMFT selected the wisest R&D path in a massive market opportunity essentially from the beginning, and now they're on the cusp of high level rewards for their wisdom. And stakeholders who perceive this with sufficient clarity in a timely manner might share in those rewards.

    All just in my personal opinions of course. And only a fool would dismiss contrary voices or strong competitors (especially Samsung) at any time.

    Highly concentrated long yada yada... (And above, oops, I meant 'carbon fiber'...)
    Jul 22, 2015. 04:03 AM | 14 Likes Like |Link to Comment
COMMENTS STATS
293 Comments
773 Likes