Seeking Alpha

H. Bruce Campbell

 
View as an RSS Feed
View H. Bruce Campbell's Comments BY TICKER:
Latest  |  Highest rated
  • IBM And Intel: The End Game For IBM Semiconductor [View article]
    Compelling Glenn. Forgive me for taking a bit of liberty please, but as you imply Watson needs the SHGB (Smoking Hairy Golf Ball). Or more precisely its modern practical incarnation, the Ball (or SB, System Ball, http://bit.ly/1pFb4Bh). My wild guess is that numerous Intel executives and top technologists have roughly 50 X scale models of SCs (System Cubes) and Balls on their desks, and are working feverishly to prepare the real things for prime time.

    With genuine heartfelt respect for most in these halls, I nonetheless feel it's occasionally very difficult for those who've amassed enormous industry experience to peer outside the industry's provincial box. Appreciation of the full implications of TSV requires a fresh view. TSV changes the stage profoundly, and my sense is that Intel fully understands this and is pursuing the technology and all its tendrils very aggressively. Coyly. But very aggressively. Because they fully comprehend the implications.

    EPhud reminds us above that Intel's quite able to pull a surprise out of their hat:

    "Intel's TriGate was a complete surprise to the industry, at least in terms of timing. No one had a clue they would have it working so soon.

    I'm still wondering if there may be some surprise there in Intel's 14nm process that hasn't been discussed yet?"

    I'm long Intel with very heavy leverage primarily for the simple reason that I believe the SC and Ball are inevitable, will spring from Intel, are an immense inflection point, and will arrive substantially sooner than most here seem to expect. (But most of my Intel calls don't expire until 2016 and 2017.)

    And I effectively concur with Glenn, and suggest that connections to near conscious processing and far more advanced search (such as search plus deep analysis) will follow the SC and Ball. First, due to processing limits, for a limited market at a rich price. Then for a broader market as Balls proliferate.

    It will take some time for those connections to reach practical fruition (though perhaps not as much time as we suspect). But my wild guess is that SC and Ball hardware are far closer than most suspect, and are wholly investment actionable.

    But I'm just speculating of course. I could be wrong. I'm often just plain wrong...

    Good luck everyone!

    PS: In my view another terrific thought provoking piece Russ. My thoughts diverge a bit at times, but you possess far superior knowledge and experience in many areas, retain a robust explorer's vision, and have my full respect. And gratitude - your insight and from the heart constructive passion has enriched my life both spiritually and materially. Substantially. For whatever it might be worth, you have my support and eternal good will.
    Sep 18 04:43 AM | 5 Likes Like |Link to Comment
  • More On Intel And Memory [View article]
    Fascinating thinking and from the heart good cheer and humor from Russ as usual - a real treat.

    I mostly defer to Russ and most others, but my instinct is to maintain my Micron call and common holdings, which dominate my portfolio, and of course my Intel (mostly 2016 Intel calls), which is my next heaviest holding, because my facts bereft instinct is that Intel's more likely to utilize Micron as a technology and fabrication resource rather than relegate them to competitor status, or even be indifferent to them.

    And I continue to believe TSV is at the core of relatively near term future drama (http://bit.ly/1pFb4Bh) - my guess is that it will become mainstream for Intel relatively soon, displacing a lot of on chip and in package technology. And my sense is that if that's the case, Intel can dominate the technology most effectively by continuing to partner with Micron. Intel might be relatively awash in fabrication resources at the moment, but technology inflection points can turn resource surpluses into resource deficits quickly. And Micron now possesses superb resources, including of course jewel class Elpida.

    Respectfully I disagree with this statement by Just Some Guy: "Current processors are already full of memory, only it's cache memory in two or three or four different levels, depending on how you look at it. It's not clear that anything else is really needed and still possible within the laws of physics."

    I own reasonably current personal IT products, and in my experience they're slow in some important respects - they seem so awash in software complexity (especially browser advertising overload) that they're frustratingly sluggish. Sadly, that software complexity overload isn't going away in the foreseeable future - rather it's more likely to get worse. Faster and more efficient hardware is the only counter. And we're simply not constrained by the laws of physics yet - further intimacy technology advancements are clearly possible, and there's enormous corporate treasure related pressure pushing them forward.

    I differ from many in my belief that TSV will prove to be the key technology for the next substantial inflection point, but by whatever means I suspect remarkable drama will unfold soon. But I suspect Intel will forge this future with little brother Micron at its side. Not significantly by virtue of affection, but simply because it's ultimately the most effective means to make life ever more difficult for Samsung, Hynix, Qualcomm, TSMC, and others.

    (Especially the Korean competitors. Those who've spent time in Korea have probably felt the rarefied atmosphere there. Neighbor North Korea imparts unique intensity to productive life - you tend to get things done with an extremely efficient war time like focus. Even if remarkably successful, Intel will, in my estimation, always feel Samsung's breath on its neck.)

    Though quite old, this (following one of Russ's leads) seems interesting in furthering understanding of some memory technology logistics, or at least the general atmospherics: http://bit.ly/1Bfbtxh.

    And a last wild speculation: Perhaps the Grantley server chip will be compatible with DDR3, DDR4, and HMC. Or, most wildly speculating, will be available as an SC (System Cube). Which, though nothing like a trivial technology advancement, conceptually at least is simply a logical extension of the existing HMC.

    Good luck everyone!
    Sep 5 12:26 AM | 9 Likes Like |Link to Comment
  • Intel And Market Action [View article]
    Thanks for sharing Russ. Your thesis seems plausible to me too (though beware everyone, I'm a relative empty shirt).

    For what it might be worth:

    The vast majority of my holdings are Micron and Intel, proportioned overall at about one third in calls and two thirds in common. But my Intel holdings are about 90% in calls - I threw the Hail Mary in Intel's case. I almost always purchase the longest term calls available, always well out of the money, and currently hold only 17 January 2015 and 15 January 2016 calls. I'm exclusively long - I don't hedge.

    I'm much heavier in Micron than Intel mostly as a matter of legacy and reluctance to divest any Micron in spite of the Intel story. I won't sell any Intel nor Micron going forward unless solid new evidence suggests I've viewed either firm inaccurately. But in recent times I've been accumulating only Intel calls.

    But my guess is that both will perform exceptionally well (Micron's warts notwithstanding). Most of you know why I feel that way, but if not it's explained at http://bit.ly/1pFb4Bh. My guess is that Intel and Micron are more deeply joined in this possible technical inflection point than is generally recognized. I also speculate that TSV development is accelerating faster at Intel than is generally recognized.

    From earlier comments, Russ appears to view my TSV, SC and Ball thesis skeptically, though I don't speak for him of course - he obviously speaks for himself superbly.

    I have no tangible idea why Micron fell so sharply today, but I view it as possible that forces similar to those Russ described are at play - perhaps large players seek to shake out worried holders so they can accumulate more Micron at bargain prices. When major players confidently perceive something big on the horizon, they finesse any market mechanism legally available to accumulate aggressively at the best prices their finessing acumen can muster. But all that's pure speculation on my part...

    Whatever the case, I hold perfectly tightly with no loss of confidence. When the TSV, SC, and Ball story become fully clear to all, one way or another, I'll reevaluate. Otherwise nobody will get their mitts on my shares or calls, no matter their ability to generate volatility. Quite the reverse, I'll continue to accumulate - especially 15 January 2016 $45 Intel calls - on dips. (But I also have an order to purchase significantly more Micron common at a bit lower than today's closing price.)

    And the Cosmos will mercilessly judge my prescience, or lack of it, later...
    Jul 28 04:51 PM | 10 Likes Like |Link to Comment
  • Micron: The Best Is Yet To Come [View article]
    Hi William. Thanks for your terrific article and very kind words.

    I can only speculate wildly about timing logistics related to TSV centric product developments. To wit: My guess is that TSV fabrication technology challenges aren't a significant element in the HMC introduction delay we're observing - I suspect the larger element is, as you reflected, 'competitive reasons'.

    Intel and Micron seem to nurture their highly constructive partnership very respectfully, so in my view there's a pretty good chance the HMC's being artificially held back in order to best position both firms for the far more significant SC and Ball drama to come. While I do think all other fabricators are perfectly well aware of the significance of the SC and Ball, and actively trying to manage the implications, I nonetheless believe it unwise for Intel and Micron to send the HMC town crier out to openly validate the matter, putting all competitors on open notice, and thus invigorating their corporate cultures. It's best to be coy, even when the elephant's obviously wandering about the room, until that elephant's completely ready to burst through the room's walls in full powerful stride, pounding the earth thunderously and braying with a deafening voice nobody can ignore no matter how much denial's in their hearts.

    My guess is that the HMC's production ready and could make some money now. But it's peanuts compared to what's at stake with clever management of SC and Ball introduction timing. If I were Intel, I'd be as coy as practical until a broad set of commodity level SC and Ball products are in full production and thus (in this case) ready to be announced. This is a dramatic technology inflection point. And wisdom dictates that it be managed in as coy a manner possible until the curtains for the gala premier event are drawn open with a flourish.

    The HMC, even if production fabrication's technically fully ready, can wait...

    And I suspect Micron fully respects this...

    However, at http://bit.ly/TsPwtt some beans were spilled in the form of an HMC enhancement to Knights Landing. This seems contrary to my 'be coy as long as possible' thesis. So maybe my thesis is just plain wrong, or needs refinement (maybe Intel seeks to publicly imply TSVs will be centric to only extreme performance systems for example, whereas in fact they're working day and night to prepare commodity SCs and Balls for abrupt introduction as my thesis suggests).

    In any case: Intensely long Micron common and calls. Heavily long Intel calls, and will continue to add on dips. Because I suspect both firms will earn a lot more money starting within very roughly two years.

    But everyone must think for themselves and steer their own ships of course...
    Jun 28 05:09 PM | 2 Likes Like |Link to Comment
  • Micron: The Best Is Yet To Come [View article]
    Phred's comments are very compelling. But I remain intensely long Micron common and calls (and remain very heavily long Intel calls), and will almost certainly hold though the earning reports as well, only slightly because:

    in my experience with a generally rising issue, when I try to trade it I lose at least as much by being left behind as I gain with successful dip anticipations. I seem to be able to trade relatively range bound issues like CEF successfully enough to justify the time, or with very small blocks (1 or 2% of total holding) of issues like Micron common during those rare periods when it's range bound due to apparent price pinning action associated with a substantial derivatives expiration date (http://bit.ly/TwklWw). But otherwise I just hold Micron and Intel common during general growth periods. And I don't try to trade derivatives under any circumstances - the fees strike me as too expensive, so I just hold them.

    But for me the far more compelling reason to hold relates to the big technical picture which I feel overwhelms the much smaller (though admittedly quite ugly) warts:

    In my view all the major fabricators recognize very clearly that TSV and related technology will define which firms dominate processor, memory, and storage markets relatively soon, and are in an all out winner take all sprint to a TSV centric product array finish line. I suspect Intel will win this sprint, and with the greatest punch, just as they have with fabrication nodes. In my estimation an ability to efficiently fabricate TSVs and stack chips like pancakes into cubes and balls is ultimately more powerful than even a multiple shrink node lead. The immense speed and power conservation benefits of SCs (System Cubes) and Balls (System Balls) (http://bit.ly/1pFb4Bh) overwhelms even multiple node advances, as the HMC (Hybrid Memory Cube) so dramatically demonstrates.

    We all know signals can travel only about the length of an index finger within a modern processor's clock cycle time, or less. And at least as importantly, signals sent off chip must be driven with a lot of power, relatively speaking, to overcome parasitic losses in the course of a chip to chip journey from one provincially connected chip to another, even within multi-chip packages. The HMC demonstrates the massive speed and power conservation gains TSV and chip stacking enables. The SC and Ball are next, and will doom the SoC to a legacy technology role. I suspect all the fabricators see this future perfectly clearly. But they simply can't outrun muscular Intel in this sprint - Intel will win this race. Because in this case Intel has not failed to perceive and aggressively pursue a critical technology inflection point. Nor has partner Micron.

    Again in my estimation all the major fabricators are intensely focused on this, and trying to determine how to survive the transition to SC and Ball technology dominance. But my sense is that by clarity of vision, technical prowess, and sheer force of size Intel will be first to market by a considerable time margin, and might own some important TSV and related IP to boot. I suspect Intel's future as a rising phoenix is superb. And Micron, as the HMC developer and Intel partner, will substantially share the fruits of this transition.

    Some disappointing daily mechanics notwithstanding (and I agree a couple of quite significant self inflicted abject waste injuries plus management awkwardness hang heavily over Micron), Intel and Micron seem to me to be on the cusp of a remarkable technology transition. My personal feeling is that they'll be first to market with the SC and Ball and will dominate the technology for a considerable time (though Samsung's a wild card). When recognition of this dramatic story begins to take root within the broader investment community, the warts will seem utterly trivial by comparison.

    My sense is that it's just too dangerous to try to trade in the vicinity of this remarkable technology inflection point, even during times when the market suddenly sees some unsightly warts - the risk of trading out and then being left behind is just too great to be offset by potential trading profits.

    Sans a looming watershed class technology inflection point, I might also go to the sidelines as Phred and others have recently, particularly in view of geopolitical world events. But for me, the TSV story is far too important to potentially miss. So I hold my precious investment jewels very tightly, ignoring bumps in the road whose timing I usually can't anticipate well anyway.

    But that's just one back woods investor's personal perspective. We'll see which view seemed more prescient later...
    Jun 17 11:34 AM | 8 Likes Like |Link to Comment
  • Micron Technology: Toxic Convertibles, Part 2 [View article]
    Oh balderdash... Such shenanigans would never so much as occur to Fred, as those who've been following him and this story are well aware. Six 'likes' for such a hostile and ill informed reaction? Yikes... It'd be a shame to see this group degenerate as so many others have...

    Thanks so much for your terrific article Fred.

    And for what it's worth, I sure miss Steve Appleton. As best I could tell, nobody held Micron closer to his heart than he did. He was very well respected and liked, and as far as I could vaguely discern matters, he ran a very clean ship. I loved the way he responded to the RamBus assault - with courageous belief in Micron (and justice).

    And though I'm strictly speculating of course, I suspect he'd have found far more efficient means to finance the Elpida acquisition (at least after the first tranche).

    The convertible bonds strike me as, essentially, bets against Micron's success. That's a bet I suspect Steve Appleton would not have allowed except when there was simply no other option. And it seems to me that there have been far better options, especially in recent times. Bets on Micron's success class options rather than bets on its failure.

    Globally investors are chasing yields almost desperately. The convertible bonds simply weren't reasonably in line with the global yield / risk market landscape - so much so that they suggest alarming questions of judgment, and questions of heartfelt stout belief in the firm.

    I miss Steve Appleton...

    (But I remain, at this time at least, intensely long Micron common and calls.)

    Thanks again for your tireless and wholly constructive energy Fred. We're in your debt.
    Jun 16 11:56 PM | 5 Likes Like |Link to Comment
  • Micron, Intel, Apple: A Tale Of 3 Companies [View article]
    Just speculating, maybe maneuvering for more economical retirement of portions of the remaining convertible debt's woven into the report in some measure.
    Apr 3 11:30 PM | 3 Likes Like |Link to Comment
  • VirnetX / Apple Appeal: Bad News For VirnetX And Other Patent Assertion Entities [View article]
    Security's clearly critical in information technology. IT firms demonstrate it overtly by incorporating highly robust and technically excellent security features in nearly every IT device. The market demanded it in every quarter, and firms responded with billions in technology development, implementation, and service investments. And IT firms continue to do so, and will as far into the future as we can see in a cat and mouse game with the unsavory side of mankind's nature.

    The market - we users - willingly pay a massive price too in personal overhead. We tolerate the inevitable high personal costs of security features which force us to manage a labyrinth of Rube Goldberg like barriers just to set up and use our own systems and their cloud companions - an often severely time wasting array of hurdles which we survive only by occasionally emitting foul language to release some of the frustrations they impose. Yet we do tolerate them - there's no significant market push-back for less security so as to relieve some of its accompanying inefficiencies and frustrations.

    There's enormous market pressure for more advanced security systems which impose less overhead and frustration, and great fortunes to be made by technology epiphanies which make them possible. No credible people are claiming fingerprint sensors are a useless fad or waste of human energy for example, and Apple's smart to be among those on the forefront of incorporating this technology. Any security related advancement which imposes significantly less overhead or more complete security, or both, is a high level technology and business treasure. Firms wouldn't fight over them like gladiators if they weren't. And the pressure for ever better security systems won't end with any new advancement - it will march on essentially forever.

    I'm far too ignorant of most other VHC factors to form views about the firm as an investment, including lack of study of the patent validity issues.

    Mike Farmwald's contribution here spawned some superbly informative debate. But a few discussion comments seem to suggest security technology's of only modest value when in fact it's critical to IT. And the money and user overhead devoted to develop and support it demonstrates this beyond rational question. Any suggestion or argument to the contrary is in my view either ignorant or intellectually dishonest. But the notion does seem to lurk in some quarters or between some lines. A wise investor should discount any such notion - it's nonsense.

    In my view any judge who fails to recognize the critical nature and sheer magnitude of security technology endangers the cause of wisely rendered justice. So I hope any judge involved cultivates a very clear understanding of the full role of security technology - if VHC's patents are valid, I hope their value will be measured with fully enlightened wisdom. And I suspect they will.

    (Positions: I hold no current position in VHC, but traded in and out last year, short term, mostly in ignorance, earning a modest profit. I may or may not become involved again, but if so only in long positions. I hold substantial long term long positions in VRNG and VRNGW, but I view Vringo as a significantly different situation. I also hold a modest long term long position in WDDD, but know very little about it - it's just an ill informed roll of minor sized dice. The vast majority of my IT time is with Apple products, but currently I hold no Apple investment position. But I have in the past, and might again in the future.)

    Oshaberi sumimasen...
    Mar 8 12:39 AM | 5 Likes Like |Link to Comment
  • Vringo's Shareholder Update And What Investors Need To Know About Patent Reform And This Patent Hero [View article]
    Judy, you might find http://bit.ly/11IsJek amusing. If not, sorry for the distraction.

    Long Vringo common and warrants.
    Feb 20 06:21 AM | 1 Like Like |Link to Comment
  • Intel: Google Is Not A Threat [View article]
    I suspect English isn't Vivek's native language. Mastering a second language is really tough of course - it's a long and daunting process, especially if begun in middle or late adulthood. It's wise to allow generous margin - mastering English is nothing like easy.

    (And even native English speakers are frequently guilty of truly hideous grammar and word misuse...)
    Dec 15 07:40 PM | 3 Likes Like |Link to Comment
  • Micron: Why I Think It's Different This Time [View article]
    Hi Thomas,

    I can only speculate about how well Intel officers visualize future options in the context of the overall technical and market landscape, but my guess is that they spend a great deal of time and energy on it, and are superbly informed and skilled at it.

    It's a remarkably complex landscape to consider. In an effort to approach a vision from the future backward, in my view all processing, memory, and storage technology is connected with ever increasing intimacy and headed squarely for a 'Smoking Hairy Golf Ball' future. Or at least a 'Ball' future. I can't imagine how it could be otherwise.

    Though some others here see it differently, I think it likely that TSV's role in this will ramp very quickly since it's such a dramatically key element in the road to the Ball. I presume fabricating TSVs is currently comparatively cumbersome, and that they currently consume a comparatively large amount of chip real estate. But multiple competitive pressures to minimize TSV fabrication and real estate costs are in my view extreme. And since no fundamental laws of nature are in the way (as far as I know), my guess is that TSV fabrication efficiencies will advance very quickly, and TSV based cubes of many kinds will become pervasive, even in very low cost end products, well before most expect.

    In my speculative view Intel perceives that almost all processors (and GPUs), memory, and storage will exit foundries in cube form in the relatively near future. Then roughly ball form as the technology becomes even more refined so as to extract every last Angstrom of size efficiency from systems - individual chips will be size configured so as to stack into final forms which are roughly ball shaped. There won't be any hair of course - connections to the ball will be far more intimate. And probably no smoke either - one of the beauties of ever tightening intimacy technology is that it yields substantially lower power requirements. Maybe liquid cooling will be required in expensive high performance applications such as enterprise markets. But my guess is that liquid gas won't be necessary in any but the most esoteric research applications (perhaps HAL's processors will smoke, but not many others).

    Intel has to at least try to perceive the future as accurately as they can. Their brass is enormously better informed and more experienced with the nuts and bolts of the technologies than I am of course, as are many here in these Seeking Alpha halls. But for what it's worth, my guess is that Intel sees a future in which processing, memory, and storage are so bloody intimate that no firm can remain competitive unless it masters all those technologies as if one technology. Because technology and competition are mercilessly driving us to an inevitable cube / ball future. So awkward or not (and it is of course frequently awkward), in my estimation Intel must plan and execute for that reality.

    So my sense is that Intel must embrace the memory and storage business again. The only reasonable way to do so is to embrace Micron much more than they already have. Whether they acquire Micron or form far more extensive and tight partnerships with them I can't judge. But if the later, I think the partnership would ultimately have to be extremely tight - a marriage in all but name.

    That's just my speculative view of course. But I'm invested accordingly, and, though perhaps for different reasons, Russ makes perfect sense to me when he states that Micron's the biggest pure play in this arena. However, already massively heavy with Micron calls, including 15 January 2016 calls, currently I'm adding to my 15 January 2016 Intel $30 call holdings during this period of Intel stock price weakness. So I hope I'm seeing things about right. And I hope the world stays reasonably glued together in the meantime, with a minimal number of only modestly sized black swans...

    Intensely long Micron common and calls, and modestly long Intel common and calls.
    Dec 4 01:49 AM | 5 Likes Like |Link to Comment
  • Micron: Why I Think It's Different This Time [View article]
    A very fresh and thought provoking look at the big picture with a clear perspective of the future, thanks so much Russ!

    In your article "Intel: Something Very Big Is Going On", you ponder the possibility that Intel might intend to jump back into the memory business, and aggressively so.

    That possibility sure makes sense to me, as I speculated at the time:

    "...If Intel calculated that they or, eventually, Samsung could manufacture storage at near enough to hard drive manufacturing costs to make whatever cost difference would exist an insufficient savings to compensate for the many dramatic performance advantages of SSDs, then they were faced with a striking new paradigm: Hard drives would essentially cease to be viable in almost all new [consumer] products.

    If that was the case, Intel knew someone would capture the new paradigm storage business at a relatively near term transition time. It seems safe to say Intel would prefer that they be the winning firm. And presumably they'd like to win the fast growing and very lucrative storage business dramatically, so that they could dominate it for the foreseeable future.

    And in my estimation TSV will play a very important role in this too.

    Just speculating very wildly, it strikes me as a bit unlikely that the US government would stand in Intel's way [of acquiring Micron]. High technology treasures are national treasures with national security implications. If the US government views the matter as a choice between Intel and a foreign firm, in my estimation they'll just wave Intel through the turnstiles..."

    As you've so consistently, elegantly, and compellingly explained so often, the memory business is fundamentally different now. So factors considered in earlier times to judge whether Intel should be in the memory business are now obsolete. It seems to me that Intel must consider the current memory business as a newly configured and extremely lucrative avenue which it's in a superb position to profit from. Beyond adding terrific numbers to their bottom line, it could also provide an important diversification hedge to insure their survival and health even if they find it difficult to dominate in guerrilla warfare like processor design competition.

    With technology advancements nearly poised to transfer a great deal of HDD money into the memory business, and Intel's enviable technical and infrastructure position, to me it's difficult to imagine that they'll not indeed jump in with both feet.

    But you didn't ponder that aspect of the landscape in this article. Second thoughts?

    Intensely long Micron common and calls, and modestly long Intel common and calls.
    Dec 3 01:45 PM | 6 Likes Like |Link to Comment
  • Is There Another Double In Micron? [View article]
    That's certainly interesting... The firm claims "...Everspin has established an MRAM intellectual property portfolio of more than 600 active patents and applications, many of which are fundamental and essential for MRAM technologies." If true to the point that other firms, including Micron, don't possess MRAM patent portfolios of comparable strength, and the technology fulfills its apparent potential and ultimately dominates, Everspin could conceivably profit substantially at significant expense to actual producers.

    The article also states "But Whether those claims actually manifest themselves in a commercial product remains to be seen, though. Commercial mass production is targeted for 2018, the report said." But the article goes on to suggest Everspin already markets an MRAM product. But I presume it's not yet genuinely competitive with legacy technology.

    As I vaguely recall, MRAM research has been active for many years, including substantial efforts within Micron. But I'm not familiar with any technical details other than a simple understanding of the allure of the technology. It would be interesting to review well informed observers opinions of the technology's potential and possible timing. Including Micron's of course...

    Does anyone recall Micron's current sense of the future of MRAM technology? Or other well respected entities?

    Intensely long Micron common and calls.
    Nov 25 05:58 PM | 1 Like Like |Link to Comment
  • Micron: Stock Is Starting To Rally Again - Time To Join The Bandwagon [View article]
    Micron's Automata Processor announcement this morning struck me completely by surprise. I had no prior awareness of this work. For me personally, it's far too early for me to try to gauge its significance. But on first glance it sure looks like it could be very exciting.

    My instinct about future computing is that it will prove to be a substantially more memory centric hardware process than in the past. That is, exceptionally efficient access to massive volumes of exceptionally well organized data will tend to define future computing and processing, particularly in key fields such as very deep simulation, such as simulation of entire living systems (such as human beings) at atomic and DNA mechanics level resolution for life sciences, and similarly exceptionally high resolution simulation efforts in physics, mathematics, and many other disciplines. And in Artificial Intelligence, which I believe will be planet defining earlier than most expect.

    The Automata Processor seems, on initial view, to target that future directly and impressively. It'll be very interesting to see how this evolves - whether it's as dramatic as the first excitement in my bones suggests.

    I don't discount the possibility of a black swan or two in Micron's future as many of you know. But as always I'm thrilled to see that Micron's doing superb technical work, and believe that will overwhelmingly define its very successful (and lucrative) future, a few warts notwithstanding.

    Intensively long Micron common and calls.
    Nov 18 06:48 AM | 3 Likes Like |Link to Comment
  • Intel: Something Very Big Is Going On [View article]
    It's a compelling concept, thank Russ!

    If Intel calculated that they or, eventually, Samsung could manufacture storage at near enough to hard drive manufacturing costs to make whatever cost difference would exist an insufficient savings to compensate for the many dramatic performance advantages of SSDs, then they were faced with a striking new paradigm: Hard drives would essentially cease to be viable in almost all new products.

    If that was the case, Intel knew someone would capture the new paradigm storage business at a relatively near term transition time. It seems safe to say Intel would prefer that they be the winning firm. And presumably they'd like to win the fast growing and very lucrative storage business dramatically, so that they could dominate it for the foreseeable future.

    And in my estimation TSV will play a very important role in this too.

    Just speculating very wildly, it strikes me as a bit unlikely that the US government would stand in Intel's way. High technology treasures are national treasures with national security implications. If the US government views the matter as a choice between Intel and a foreign firm, in my estimation they'll just wave Intel through the turnstiles.

    Intensely long Micron common and calls, and modestly long Intel calls.
    Nov 15 04:50 AM | 2 Likes Like |Link to Comment
COMMENTS STATS
188 Comments
206 Likes