Seeking Alpha

Justin Jaynes

View as an RSS Feed
View Justin Jaynes' Comments BY TICKER:
Latest  |  Highest rated
  • Thursday Will Likely Center Around PCs For AMD [View article]
    hey stocks -

    regarding supply chain type stuff, my *guess* is that we're talking about a 1 to 2 month lag time between when AMD ships a chip for a product launch and when that product comes to market, but that's just a rough guess on my part.

    so for products launching in late June (there were a few laptops announced yesterday), amd likely started supplying those chips after march, so there probably isn't much of an impact to Q1 revs.

    again, this is only a rough guess on my part, and it could be off
    Apr 16 05:08 PM | Likes Like |Link to Comment
  • Why AMD Is A Good Short To $3.00 [View article]
    What does the number of articles you are putting out have to do with factually incorrect information?
    Apr 16 01:11 PM | 14 Likes Like |Link to Comment
  • Why AMD Is A Good Short To $3.00 [View article]
    "What would be the point?"

    The point is to put out factually correct and useful information

    -AMD has 2 products at 32nm, and most are at 28nm.

    -Console profitability has definitely went to the bottom line. Q3 2013 was the first time AMD was profitable since Q2 2012, and this was due to the console win.

    -Your article is a lot of opinion with little source material. For example, some links to help the reader understand how tablets will eat into console sales, and when you expect this timeframe to play out, since console sales are outpacing expectations.

    -Also, you call the console cycle "unpredictable." What makes you call it "unpredictable", and why are you making the case that console sales will fall short of expectations, which is going against the trend, rather than exceeding expectations?

    These things would help the reader understand your rationale and make the article seem less of an opinion piece and more useful to the reader.

    Long $AMD
    Apr 16 12:54 PM | 24 Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    "Building a new business takes time: years, not weeks."

    Exactly. And during these "years" this business is building (my timeframe for HSA to play out is at least 18 months at a minimum, and by play out I mean get to where it could actually generate revenue - just my opinion on the time frame, 18 months to a few years) AMD has to do things like keep the lights on and pay employs. This comes from operations, and AMD is in the business of selling chips.

    So it doesn't matter that AMD has the best HSA APU on the planet right now, as there's basically no software support for these features. Until there is, these chips have to stand on their own merits to fund the operations of AMD.

    Until HSA becomes useful, it's not going to drive chip sales.
    Apr 16 12:19 AM | Likes Like |Link to Comment
  • Windows XP's Death Will Breathe Life Into Microsoft, Intel And AMD Shares [View article]
    agreed - from what I have seen (speaking anecdotally, only referring to my office), we finally switched from XP to Win 7, not 8. I would rather shave myself with a cheese grater than use windows 8.

    Disclosure: I own multiple PCs with various OS's, and hate win 8
    Apr 15 10:24 PM | 2 Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    "AMD’s newest, highest performance APU processor"

    If it's shipping in the beginning of May, I might get me a new toy sometime soon :D

    All speculation at this point - but really hoping to hear "volume shipments of mobile Kaveri" on Thursday
    Apr 15 08:30 PM | 2 Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    Note to Andreas Hopf: You Should've Bought Cheeseburgers
    Apr 15 08:08 PM | 2 Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    @Rand - seems like you're creating a straw man with this one to beat up on.

    You are reading way too far into what I'm saying - I'm saying that AMD has good graphics IP, and compared it to an existing solution. You turn this into an HSA discussion; something I did not bring into the equation.

    More people likely care about gaming than HSA, so the market is bigger.

    Everything gets turned back into an HSA discussion with you - so let me ask you this:
    1) When do you think the timeframe for this to become a valuable technology to AMD is?
    2) What is the likelyhood you expect it to succeed?
    3) How much do you think HSA will actually be worth to AMD from a revenue stand point?
    4) What are the challenges to adoption?

    HSA is a long term bet, as software isn't ready and development tools are just now being released to the public. AMD is the only vendor shipping HSA hardware, with Imagination Tech being the only other having been on the record, as far as I'm aware, stating they have plans to ship HSA hardware, but no time frame is given.

    Apple released OpenCL in 2008 maybe? It's 2014 and OpenCL adoption is just now becoming wide spread. Not to mention Apple is worth around half a trillion dollars and has plenty of money to put into something like this.

    Contrast this with HSA, which, almost 2 years after the founding of the HSA foundation, still hasn't ratified the HSA 1.0 spec yet.

    Regardless of whether you're a believer in HSA or not, it will take time to play out, and it's not a guarantee because AMD is the only entity with HSA hardware out. If the 1.0 spec were out and the ARM camp were releasing lots of hardware, then this could likely drive revenues a little faster. ARM, during AMD's APU 2013 summit, basically stated that they didn't really see the use for ARM in small systems or HPC.

    "HSA is good for midrange systems, but I'm not sure if it scales to the smallest systems, and I don't think it does scale to high-performance computing with tens of thousands of processors," he said during the panel discussion. "One system's architecture won't scale across the whole space."

    Meanwhile, AMD has a gaming strategy - the gaming market is a huge market, and the PC market is 2x times or so larger than the server market and GPU market combined, and AMD has great graphics IP, and their low powered processors are more powerful than Intels, and the graphics can't really be matched by Intel at the same price point *OR* power budget.

    Saying AMD should take advantage of their current strength and apply it to a market that is several times larger than the one that HSA may be important to in three or four years (or whatever time frame you want to put on HSA to make it a relavent technology) is inline with AMD's gaming strategy and aligns to the its strengths - gaming and graphics.
    Apr 15 05:23 PM | 1 Like Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    thanks for the back-up, that was poorly worded. Essentially, I was pointing out that Q2 ends in June, so if Q1 products are launching late in Q2, those products that are launching late in Q2 will likely not impact Q1 revenues very much.

    With computex in June, it's likely we'll see a lot of these platforms launch in the june timeframe, prior to "back to school" season, but launches this late probably won't affect Q1 very much.
    Apr 15 04:18 PM | Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    thanks much AE! =]
    Apr 15 04:04 PM | Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    thanks tah - specifically, I would like to see, to make up a word, "mid-end" cat core APU with 4 Puma cores clocked 2 GHz+, 4 GCN CUs around 720 MHz, and a dual channel IMC with support for up to LPDDR3.

    This would probably fit in a 35W budget, and according to chipworks teardowns of a 7970 I believe (don't have the link handy), a GCN CU is about 5 mm^2.

    So AMD could field a chip that's almost half the size of kaveri (probably around 140 mm^2 compared to Kaveri's 245 mm^2), it would best any Intel HD4000, likely HD4200/4400/4600 as well, in a small enough die size to be sold cheaply and profitably.

    It'd probably offer 80% of Kaveri's performance in 50-60% of the die area, and you could tweak clock speeds a little to offer a few SKUs down into the 20W-30W range.

    Products like this could then be marketed under the "If it can game" campaign.
    Apr 15 02:27 PM | 3 Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    thanks stock -

    there are so many moving pieces and parts to AMD right now it's really hard to know where the chips will fall.

    For example, extreme demand for GPUs didn't really help much in Q4, but AMD was producing damn near everything at TSMC and the consoles are massive chips. AMD expected console demand to fall off this quarter, so they probably could've allocated some more capacity to GPUs in Q1, but we know that high prices for GPUs persisted for most of Q1, so it's hard to tell how much AMD benefited, as the supply was the issue during Q4, and AMD doesn't give a lot of granularity on supply.

    Sony can't make consoles fast enough, and XB1 seems to be selling reasonably well.

    So biggest unknowns - is AMD still hanging on to desktop market share? When will all these mobile platforms launch, and when will they impact revenues?
    Apr 15 02:04 PM | Likes Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    thanks toonies, avatar, et al for the positive feedback.

    i know I'm usually in the right ball park if I can write something that doesn't garner a lot of comments from either side of the fence.
    Apr 15 01:55 PM | 1 Like Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]
    I always focus on performance of AMD's products because there aren't very good design wins featuring AMD CPUs and APUs, so only those that really understand AMD's performance get the value - this refers mostly to the CPU space rather than the GPU space.

    If AMD had better design wins or better brand recognition, there market share would likely be much higher.

    And I'm somewhat critical of AMD's current "If it can game" market strategy, because the company is trying to draw parallels between the console chips and AMD's standard product line. The issue here is that there are no "halo" products to make this marketing strategy really make sense.

    However, at least over the last year they have taken a specific direction in marketing. AMD always talks about "showing customers the benchmarks", and I think it shows with product releases like the AM1 platform having widespread support at launch with numerous motherboards. The "If it can game" strategy still seems dubious, as there's really no wow factor to it, and it's not useful to AMD's core consumer.

    AMD needs something that speaks to the core consumer but is understood by everyone.
    Apr 15 12:06 PM | 1 Like Like |Link to Comment
  • Thursday Will Likely Center Around PCs For AMD [View article]

    That's essentially correct, but there is a subtle nuance to understand here.

    Most gaming consoles in the history of gaming have been somewhat proprietary designs that were seperate from tradtional PCs. Most consoles have either used ARM processors (IIRC, in the case of hand helds) or powerPC or ARM in the traditional consoles.

    Using the PS3 as an example, Sony worked with various others to develop the Cell CPU, and hired Nvidia to build a proprietary GPU that could only be used by Sony. The result was that Sony had incredibly expensive, powerful hardware thas was difficult to work with and took a bath from each console sold - they lost money on console sales with the goal of making it up in software sells and licensing.

    Fast forward to now, and gaming hardware designed for the PC is extremely powerful and volume economics are working in advantage for the PC.

    So with the PS4 and XB1, AMD said "hey, I have these x86 cores and these graphics blocks you can use, what do you want me to build?"

    Then AMD worked with the manufacturers, using IP from GPU and CPU/APU operations, and combined this IP with requests and IP from Sony and Microsoft to build the console APUs.

    So the Jaguar blocks and GCN cores used in the console APUs come from traditional operations, meaning AMD had to spend very little money on this R&D. Sony and Microsoft help pay for NRE (non recurring engineering expenses), so part of the R&D is paid for, but not all.

    Semi-custom works on the premise that AMD has IP that is largely paid for in other areas of operations. This, combined with minimal marketing and admin expensive, and the customer paying for the NRE, allows AMD to increase operating margins at the cost of gross margins.

    Sony and MSFT get a great deal, AMD gets a contract giving them a guaranteed design win, and a small amount of profit per console, but in very sizable volume.
    Apr 15 11:24 AM | 3 Likes Like |Link to Comment