Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Isn't that a rumor that you started?
What's amazing to me is that AMD management held a dog-and-pony show and loudly declared that GMs would be in the 55% range. Only one month later GMs are at 36% (note that the 40% figure is non-GAAP, although even if we give them the benefit of the doubt and use 40%, it's still pathetic coming right after that analyst pump-n-dump meeting).
Chipguy, I think you were exactly right when you previously said that AMD's purchase of ATI was done at a time that would best hide the troubles coming ahead. The first headline that came up on Yahoo says, "AMD income down on merger costs". Never mind the real story:
-----------------
Change
------------------------------------------
Q4-06 vs Q4-06 vs
($M except percentages) Q4-06 Q3-06 Q4-05(2) Q3-06 Q4-05
----------------------------------------------------------------------
Revenue
----------------------------------------------------------------------
Pre-acquisition AMD $1,374 $1,328 $1,351 3% 2%
----------------------------------------------------------------------
----------------------------------------------------------------------
----------------------------------------------------------------------
Operating Income (Loss)
----------------------------------------------------------------------
Pre-acquisition AMD 63 142 272 (56%) (77%)
----------------------------------------------------------------------
I find the last sentence below to be very significant. Remember that the server Woodcrest part was the first NGMA product released back in June, and has had a full 6 months to ramp.
I think next quarter, we could start to see a similar story for desktop, and possibly even mobile (though I think they will be able to ride the notebook craze and continue posting gains compared to their tiny starting point).
Fourth quarter demand for AMD mobile processors was especially strong, resulting in record unit shipments and revenue. Mobile processor unit shipments and revenue both increased 41 percent quarter-over-quarter. Year-over-year, mobile processor unit shipments increased 76 percent and revenue increased 85 percent. Desktop processor revenue was also strong in the quarter, led by demand for AMD Athlon(TM) 64 X2 dual-core processors. Overall server processor unit shipments were essentially flat compared to the third quarter and ASPs were down significantly.
Projecting a Gross Margin of 50% for all of 2007 is very....underwhelming. This is the key sentence in that entire earnings report, IMO:
Gross margin included the positive impact of higher microprocessor units and selling prices that were partially offset by higher factory underutilization charges along with flash memory write-downs and NAND start-up costs.
This statement contains both incredibly good news, but also disheartening news IMO. I think we all expected unit shipments to be higher, but what should be a home run is the fact that ASPs not only remained stable, but went UP. That's a giant middle finger to all those analysts and 'droids who think that this presumed "price war" is dragging down Intel along with AMD (in contrast, this shows there is merely a "price battle" between Intel's low-end chips and AMD's entire product line, while Intel's new higher-priced offerings remain above the fray, much like we've all predicted here for 6 months).
Unfortunately, there is apparently no financial reward for this outstanding execution, because Intel evidently has too much capacity, and flash continues to drag everything down with it. While I am thrilled to see how Intel is responding to the competitive threat of AMD, I guess I would like to see just a bit better basic management of internal operations.
If GMs for the year really do turn out to be 50%, then I'm afraid that this stock will go pretty much nowhere this year. It's possible that management is sand-bagging and projecting the worst-case scenario, but I've played that game for the last 2 quarters, and am getting a little weary of it...
While year-over-year comparisons should improve in 2007, I have yet to see any signs that even minor growth can be maintained beyond that.
As an INTC stockholder, color me disappointed...
Of course, now the big question is, sell the puts and take the 50+% gain we got today, or hold through earnings?
Tough call...my price target was $18 in Q4, and $13-15 by Q3. They've hit my price target for my April puts, so discipline says I should sell, but I can't shake this feeling that things will be even worse once they factor in the ATI results.
We probably shouldn't underestimate AMD's ability to spin things at their conference call....
I read on a newsgroup that the iPhone processor is from a company called PortalPlayer, who I think Nvidia recently acquired. Don't know any more than that (or even if it's true). Jobs said in his presentation he has been looking forward to this product for two and a half years, so its development predates the Intel relationship by quite a bit. Maybe the next-generation product in a few years?
Confirmed that Intel flash is used, though.
Oh, and the iTV that was announced uses an Intel processor (apparently a Pentium-class, though I'm not sure which). That product doesn't look to be that big of a deal, though. The iPhone looks for real, although it sure is spendy...
I just think of Mr. Market and all the INTC analysts as playing the role of Cuba Gooding Jr. on one end of the phone jumping up and down and yelling, "SHOW ME THE MONEY!"
Until they see increased profits and increased forward guidance, we'll continue to see wild speculation that nobody really listens to, and INTC stock will continue to go absolutely nowhere. Fortunately, I'm a patient guy.
Until INTC takes off, however, there is plenty of money to be made riding AMD down to the lower teens over the next year...
Re: AMD puts
I'm still holding on to the April 22.5 puts that I bought just prior to the Q3 earnings announcement. After earnings, I was sitting on a nice 50% gain in one week, but decided to hang on through another quarter (I regretted selling my last puts too soon over the summer).
I was regretting that decision a few weeks ago, when AMD's analyst day brought me back to just about break-even. But with today's action, I'm back to about a 60% gain, and am feeling confident heading into the Q1 release.
I will probably sell these after earnings regardless of what happens, because the analyst day told me that AMD is going to be going into full denial and dream mode, and although this should make AMD's looming drop that much more precipitous, it makes it very cloudy as to WHEN this will occur (the market is just full of suckers). If the stock gets back up to the 21-22 range, I very well may get back into this position with some longer-term put LEAPs so I don't sweat about it quite so much. Less potential gain, but more of a sure thing, I think...
Here is the exact text of the "Future Plans" column from the link you provided:
D1C: "65 nanometer in late 2006"
D1D: "45 nanometer in 2007"
You are assuming that these 2 events are 1 year apart. If they were really 1 year apart, wouldn't D1D have specified, "late 2007"?
No, they didn't say that, because D1D is focused on 45nm RIGHT NOW. As beamer pointed out, Penryn is the lead 45nm vehicle, and it taped out weeks ago. There will continue to be steppings, dash steppings, and other process tweaks for a full year before 45nm Penryn products are actually sold.
You are the one who is trying to make the number 3 be something besides 3. Good luck with that, but you're not convincing anyone here....
I am going to jump to the conclusion that Barcelona clock and performance will be crappy at introduction.
Careful there. This conclusion can only be jumped to if the lackluster results we're seeing with this first 65nm product is really because of problems with AMD's 65nm process. If that's the case, then I agree it would be logical to extend similar lackluster expectations to Barcelona and future 65nm chips.
The problem is, we just don't know if there's an issue with the process, or an issue with the design. This is a trap that AMDroids fell into when Intel released Prescott. There was no clock-speed benefit, there was no performance benefit, and there was no power benefit. Everybody was convinced that the manufacturing giant had finally fallen, and their 90nm process was crap. What many didn't realize, however, is that Prescott was quite simply a very bad design that was mismanaged from the start.
How can we tell if the issues with Brisbane are because of the process or the design? We can't. What we can do, is apply some logic to how AMD might have treated this design, and I can think of two issues that could impact the success of this 65nm lead vehicle:
1) AMD's engineering resources are likely limited because of the full-court press to get Barcelona out the door, and there was probably only a small team available to do this shrink. So I would expect not many optimizations to be done to the design to increase frequency.
2) With Barcelona around the corner, the expected lifetime of this stop-gap product isn't very long, which would further likely limit the effort they would want to put into optimizations.
Barcelona won't have these issues working against them, and I will expect to see the best effort put forth into the design of that project. Speaking personally, I am not yet ready to extrapolate the results of Brisbane to any future 65nm product from AMD.
At least somebody seems to be getting it:
http://biz.yahoo.com/seekingalpha/061214/22474_id.html?.v=1
Yesterday, Sumit Dhanda, an analyst with Bank of America, dropped his price target on AMD to $20 from $23, citing an erosion of average selling prices. The call was timed a day ahead of today’s AMD meeting with analysts.
So, clearly, Dhanda is not having a good day: the stock is spiking, apparently due to the aforementioned disclosure that it will have much greater than expected cost savings from its recently completed acquisition of graphics chip maker ATI Technologies.
Is Dhanda backing down? Nope. In fact, he just whipped out a note asserting that the spike in shares today is unwarranted. Dhanda contends that the rally is likely to be short-lived, since “consensus estimates for ‘07 for AMD-ATI will likely need to be reduced further, given that AMD has just reset its ‘07 [gross margins] and implied operating margins below prior forecast.” He adds that “new details on depreciation and tax rate imply even more cost headwinds” compared with prior projections.
The company today said it expects gross margin to be 50%, plus-or-minus 2%; he says the the consensus estimate assumed a gross margin higher than 50%. He also notes that the company now sees $1.3 billion in depreciation in 2006, up from about $800 million in 2006, and above his previous estimate of $1.07 billion.
Dhanda also notes that AMD had previously said that it expected a post-merger operating margin of 18%-24%; but he says that today’s announced targets for R&D (19% of revenue plus-or-minus 2 point) and SG&A (16% plus-or-minus 2%) imply operating margins of 15%.
Dhanda asserts that the higher cost synergies from the merger “has much to do with the fact that ATI’s revenues are trending meaningfully lower than at the time of the merger announcement.”
Finally, he notes that the expected tax rate of 26% is above his prior estimate of 20%. And he points out that the company guided to negative $500 million in free cash flow.
“Net net, a disappointing financial update, yet not too surprising,” he concludes. “We believe the margin reset…[is] likely to be a disappointment for investors. The fact that AMD did not trim its near-term targets means nothing to the stock, as consensus ‘07 EPS estimates will need to be reset lower.”
AMD today is up 11.9%, to $22.57.
More bullish remarks from AMD. Does one bet for them, or against them?
I believe them when they say that they will grow unit sales at a faster pace than the industry. Will they get to 20% unit growth? Perhaps...
What they have failed to mention is at what cost that unit growth will come.
Oh wait, I see a commont on gross margins in another article:
http://yahoo.reuters.com/news/articlehybrid.aspx?storyID=urn:newsml:reuters.com:20061214:MTFH99609_2...
Chief Financial Officer Robert Rivet told analysts on Thursday that he expects AMD's 2007 gross margin to be 50 percent "plus or minus" two points. The long-term target is for gross margins to be in the 51 percent to 56 percent range, he added.
Now THIS is where we start to question what they're saying, IMO. With immense competitive pressure from Intel, and the integration of a lower-margin business, the above statement seems like a pipe dream to me. Not to mention all those interest payments they now need to make. But I haven't run any numbers yet; anybody have any insight? Sarmad, how do your previous acquisition calculations get altered given the increased cost synergies with ATI that AMD has announced?
Elmer, rather than think of what might have been had it not been for a noise problem in the MTH, I think it's more appropriate to speculate on what might have been had Intel not gone to bed with Rambus, and made the boneheaded decision to pair that expensive, power-hungry memory with a CPU designed for low-cost and small form factor. That strategic decision still boggles my mind today, and whoever made that call (I'm assuming it's Yu), deserves all the criticism they get.
As for Pentium 4, it really is a shame that P4's legacy will be defined by Prescott. One little known fact is that the only reason Willamette (and by extension, Northwood) ever became a viable product was because it actually came in about 25% BELOW its power budget. The pre-silicon timing model was set at 900MHz, and the expectation was that silicon would hit about 1.2GHz or so (which it did). Unfortunately, one thing that became clear late in the design, as well as on first silicon, was that the IPC hit of Willamette's uarch was quite a bit bigger than expected (and not just because of the longer pipeline, but because of other less-understood features like data speculation and replay that are harder to model in performance simulations).
As you know, Willamette debuted at 1.4 and 1.5GHz, well above its initial target. The only reason this was possible was because the power was so far below initial targets, and voltage could be increased without any impact to the platform. There was just a huge level of paranoia about power consumption throughout the entire project. Nobody will ever know the extreme extent to which clock-gating was applied on Willamette and Northwood. I remember my manager sitting me down a full 2 years before tapeout, and giving me the message that if we didn't take steps to curtail power, we wouldn't have a product. We all knew we were taking a quantum leap in frequency, and a culture of power-awareness was fostered from day one.
While Willamette wasn't much of a product because it was on an older process, we all know that Northwood kicked some AMD ass in its day.
Now fast-forward to the planning of Prescott. Evidently, the people in charge (again, I'm assuming it's Yu, this was back in 2000 or 20001) didn't bother to try and take any learnings away from the Willamette design. Instead of trying to correct any IPC shortcomings (even if it comes at a hit to frequency), they mandated an opposite direction: lengthen the pipeline to get a 20% frequency boost on the same process, and when we go to the next process....WOW, look at those frequency numbers! Yippee! Oh, and since power was never an issue on Northwood, we won't need to pay much attention to that, will we? Ok folks, get to work!
Obviously, the result was a disaster. I haven't even gotten into other planning blunders like doing a brand new design on a brand new process using a brand new design methodology and a brand new circuit topology, all with the schedule of a normal compaction. Or having a split design team in California and Oregon, or having 3 project co-managers where nobody is in charge, etc, etc. The seeds for this failure were planted back in the '00-'01 timeframe, and while there is plenty of blame to go around, Yu and Barrett certainly deserve a fair share of it.
The good news is that nobody who had anything to do with planning or leading that project are to be found in any place of any significance, as far as I can tell. And the other good news is that while the Intel Israel team gets all the mad props for being so power-savvy, the P4 design team in Oregon is no slouch when it comes to managing power consumption, if only they can get a decent uarch to work with. *coughnehalemcough*
And of course, there's still a third item that I haven't mentioned. I assumed an almost unheard of level of validation time, in that AMD would be able to validate and debug their parts all within a month of silicon.
Yes, absolutely.
In my mind, the only hope of a Q2 launch is if B-0 did in fact tape out a couple months ago, and if that full stepping is completely healthy; it will take until the end of Q2 to qualify such a part for full production. So I agree with your best-case scenario, that a healthy B-0 is AMD's only hope of meeting Q2, and that's only possible if that design is already done. If they need to do any more work to productize Barcelona, we're looking at Q3 or even Q4. Realistically speaking, I think a productized B-1 stepping is the best anybody can hope for, and that's in Q3.
It speaks volumes to me that AMD management is making their external commit-date as the perfect-case scenario. Smart companies set a reasonable external commit that takes into account possible contingencies, and then shoot for an internal target to try and pull the schedule in. Gee, I wonder companies do that? *coughintelcough*
First of all, regarding the 3 months between A0 tapeout and Charlie seeing actual wafers, I'm not willing to give anybody enough credit to think that AMD would include a journali....er, whatever Charlie is, upon immediate availability. In fact, in the 3rd article you linked on October 27th, Charlies says:
"AMD HAS HAD quad core parts for weeks before we published this."
Where "this" is a link to the September 29th article that you reference. Based on my experience on past projects, a high-priority lead product should be able to get pushed through in less than 3 months (I come from the design side, though, not the fab side). Of course, it's certainly within the realm of possibility that a bug was found pre-silicon that warranted another tapeout before A-1 came back. I wouldn't call this desparation, though; they may have just realized that the problem would make A-0 more-or-less worthless to debug, so they decided not to wait. Wouldn't be the first time that's happened. Like I said, whatever the case, two dash steppings indicate to me that first silicon just wasn't very healthy.
Regarding my A-1/A-2 mixup, sorry, I was pressed for time and did too much skimming and not enough reading. After reviewing your timeline again, I agree that the existance of an A-2 tapeout in October would preclude any sort of a parallel B-0 effort (otherwise they would have just done a B-0 instead of A-2). However, do we know for sure that AMD really did do an A-2 stepping in October, and not a full B-0 stepping? The only reference I can find is Charlie mentioning "another batch in the oven", and calling it an A-2, which could be pure speculation on his part. So I think it's still possible that a full B-step could have already taped out.
I'm clinging to this notion, because I think we can all agree that if they don't yet have healthy silicon, and B-0 hasn't even taped out yet, that a product launch in Q2 is a pipe-dream. I know, I know, it's possible that AMD management is doing a pump-and-dump; I guess I don't think they would be THAT glaringly stupid, and I put more faith in Charlie getting a technical detail wrong...
I still think a Q2 launch is a huge long-shot, but it's still possible if the last tapeout was a more healthy B-0.
Nice analysis.
I think your 3-months fab time for A0 is probably conservative. You've got to think that their brand-new 65nm design would get expedited. So I wouldn't immediately assume that they taped out A1 before getting A0 silicon back, although regardless of the circumstances, doing 2 dash steppings would point to relatively unhealthy A-step silicon (especially since they apparently still don't have it right even with A2).
There is one other possibility....they may have taped out B0 some time ago, shortly after getting A1 back (or even A0, although I doubt it). In parallel, they could have done some less-robust metal-only fixes for an A-2, whose only purpose was to provide a demo before end-of-year like they promised. Having B-0 tape out back in September or October is the only scenario I can see where a Q2 launch even becomes remotely possible.
AMD has to know that everybody is anxiously waiting for their response, and if I'm AMD and I have an un-demo-able A-step come back from the fab, I can totally see doing another dash-stepping whose only purpose is for a demo.
As an INTC investor, I have to assume the worst, and that B-0 taped out a month or two ago. Conversely, if I were contemplating an investment in AMD, I certainly wouldn't bet on that scenario...
All they need is to sell a thousand of them. There may be a market for that. But outside of that handfull of chumps, the rest of the world is going to look at this and project AMD as a company losing their leadership position, and desperate for a response to their much stronger competitor.
And yet, AMD is apparently rewarded by the clueless media for this nonsense. Their shares are now up over 2.2% on the day. The amount of momentum these guys have is just insane.
http://biz.yahoo.com/ap/061130/sector_snap_semiconductors.html?.v=1
NEW YORK (AP) -- As semiconductor stocks edged higher in Thursday's trading, shares of chip maker Advanced Micro Devices Inc. got a slight boost after it unveiled a new processor platform meant to compete with an offering from rival Intel Corp.
On Thursday AMD unveiled its Quad FX Platform architecture, a desktop computer processor platform targeted to gamers and power computer users. The platform is powered by pairs of dual-core processors.
AMD shares were up 37 cents, or 1.8 percent, at $21.71 in afternoon trading on the New York Stock Exchange, while Intel shares were up 3 cents at $21.27 on the Nasdaq.
Well, there are customers, and then there are OEMs. The OEMs would love Intel and AMD to be on equal footing, so they can play them off each other and get the lowest prices.
But the customers will want the best product. And if the customers are demanding way more Intel products than AMD products, then the OEMs will be forced to comply.
This is what I think will happen over the next 3 quarters.
I wanted to collect my gains while I had them.
Yep, if I'd had Jan options, I'd have done the same thing. I was expecting a bounce from yesterday, but am pleasantly surprised to see the downward momentum continue. I won't be surprised to see some recovery, but I'll remain confident that I'll get my payday after Q4 results/Q1 outlook (and maybe even after a Q4 warning in mid-late December).
I just don't think AMD management can pull off what they're claiming. 55-60% margins? You've gotta be kidding me....
Thanks! I sold my AMD Oct. 22.5's this AM for $1.10
Well, I guess I'll just have to go it alone without you and beamer. I bought my April puts with a specific goal of going through at least 2 earnings seasons, and I'm staying put (pun intended). I still think this is a $15 stock in 2007, and my price target for these puts is $18 by April. That should get me about a 150% gain. I'm up 60% so far...
Ok, I just saw what wbmw was talking about....a $20 million tax benefit on top of $131 million in operating income? WTF is that about?
These numbers stand out:
Q3 2005 results minus proceeds from the spun-off memory division that quarter:
Revenue: $1.01 billion
Op. Income: $129 million
Q3 2006 results:
Revenue: $1.33 billion
Op. Income: $119 million
Revenue increased 32%, but all those extra sales actually decreased the bottom line by 8%.
Maybe the report isn't quite as rosy as I initially thought...
Does my heart proud to see dear old Uncle Sam support a publically traded company using my tax dollars (NOT). If anybody finds out what's up with that, I'm damn curious to know...
Here is the info on ASPs from their press release:
"Demand for AMD Turion(TM) 64 mobile processors was especially strong, resulting in record mobile processor sales and unit shipments coupled with increased average selling prices (ASPs). Record AMD Opteron(TM) processor sales resulted from continued adoption of dual core processors, record unit shipments and improved ASPs.
Desktop processor sales were flat sequentially with increased unit shipments offset by decreased ASPs."
So to summarize:
Mobile (Turion): record units, "improved" ASPs
Server (Opteron): record units, "improved" ASPs
Desktop (Athlon): increased units, decreased ASPs
The chinks in the armor are appearing in the desktop segment, but it looks like AMD has a good head of steam in mobile and server that will probably take time to overcome.
I just don't understand how Turion can continue making progress in both units and ASPs. Then again, Merom was just released in August.
I was starting to get my hopes up, but I think it's back to "wait 'til next quarter" mode for any kind of major change in perception to occur with Wall Street analysts with regard to Intel and AMD.
Re: .27 ex items
Actually, it's 0.27 excluding stock-based compensation, which they are still mentioning to give an accurate comparison to last year.
The one-time items actually contribute to earnings:
"Operating results included a gain of approximately $100 million from the sale of a portion of an investment in Micron Technology, gains from divestitures of approximately $130 million and a restructuring charge of $98 million. These items resulted in a net positive impact to EPS of approximately 1.5 cents."
So excluding one-time charges, EPS is more like 20.5 cents. That's how I'm reading it, anyway. Still good, but the inventory writedown kept it from being a knockout quarter, IMO.
http://www.intel.com/intel/finance/earnings/IntelQ32006EarningsRelease.pdf
If I remember right, you bought those INTC calls right before earnings, and they pretty much became worthless after the earnings report was underwhelming. I thought that was a crazy move at the time, but it obviously worked out in the end.
I think the AMD puts could work out the same way. I expect AMD to be in full-on pumping mode during this earnings announcement, and I think there's a decent chance the market could buy it. So things could get ugly in the short term.
But one catalyst I expect to make noise is the quad-core intro in November. I think we could see some serious mindshare gains, and the media angle will probably start to trumpet all the gains that Intel has made at AMD's expense this year (it's getting to be a pretty long list now).
I'm convinced AMD is a $15 stock. We probably won't get there this year, but I think the journey starts in Q4.
I'm right with ya, Beamer...
I actually bought the April 22.5's. A little less cash outlay with a little more risk, but I think this is offset by being able to hold these through both Q3 and Q4 earnings.
Good luck!
- CG
I am quite clueless on the technical aspects of CSI. Does anybody know if CSI will somehow prevent Intel from easily creating these MCM dual-core modules once they transition over to this new bus topology? Will it be easier or harder to do this than it is with an FSB topology?
The reason I ask is this: It's no secret that Intel's first MCM dual-core chip, Smithfield, was a day late and a dollar short, and likely contributed to this bad rap that the MCM approach is now getting.
But when I look at the situation today and extrapolate out into the future, I see a very big competitive advantage to this approach. Less than 6 months after introducing a "native" dual-core chip, Intel will roll out a quad-core offering, months before AMD can.
Of course, AMD has a "native" quad-core chip in development, but so does Intel (at the 45nm node). AMD's will come out first in mid-late 2007, but when Intel's is released, can't they just slap two of 'em together and push the market ahead to 8-core right off the bat? Wouldn't this in effect be YEARS before AMD could do anything "native"?
My prediction: before this decade is out, we will have fully transitioned from a "GHz war" to a "core war", and will see AMD copy this MCM strategy just to keep up with Intel.
The one hangup to this is my complete lack of understanding of how CSI works, and whether an MCM module is feasible using CSI.
Intel's Core 2 Duo TV ads:
http://www.youtube.com/IntelVlog
Now is about the right time to buy some AMD puts. What do you think?
I'm feeling a little gunshy on AMD puts with all the positive stock momentum. I'd rather wait for a downtrend to begin, and if I have to wait until after earnings for that to happen, so be it. The 25-26 range was my target for buying puts a month or so ago, but when it shot up there so soon and then on past 27, I decided to wait and see where it might top. I can always buy those puts when the stock is at 25-26 on the way back down...
Finally, what you call a "trend", I call a 3 year Opteron spike and a 6 year Intel hiatus from listening to its customers and losing focus on its "prime directive".
This is the key, IMO. While it's easy to say that AMD is stronger at the end of this competitive cycle than they were at the end of the last one, the main difference this time is the extra length of the cycle that has allowed AMD to further advance its position than it might have in the past.
If Intel were doing things right, Prescott would have advanced Intel's competitive position over Opteron, and the next cycle of Intel dominance would have started 2 years ago.
But Intel fell flat on its face, and the spent the last 2 years executing a right-hand turn, all the while allowing AMD to continue its steady rise up to its current point.
I'm under no delusions that the tide will turn back towards Intel quickly; but it WILL turn. It will just take longer to get there, since AMD had extra time to extend this current cycle.
It's possible AMD's bottom might not be as low as it was in the past, but I guess we'll find that out in a couple years...
I'd say they're already pretty close, Paul Otellini is already on Google's board.
http://investor.google.com/board.html
Of particular significance to me is that this was included in the very first revision of the erratta document.
In other words, the part was PRQ'd (approved for initial sale) with the full knowledge of this "problem". I have known several instances where Intel has delayed important product launches because of relatively minor issues. The fact that this made it through the final qual review is enough to tell me that this is a non-issue.
Still, it wouldn't hurt to contact either Intel or HP to see what's been done (if anything) about this. I'm not sure what good it does asking for guarantees from an investors' message board...
Here's the more "complete" description (I use that term loosely ;p):
"When request for data from Core 1 results in a L1 cache miss, the request is sent to the L2 cache. If this request hits a modified line in the L1 data cache of Core 2, certain internal conditions may cause incorrect data to be returned to the Core 1."
If this problem made it through pre-silicon validation, plus 2 full steppings and a dash-stepping before being found and/or fixed, I would bet that the "certain internal conditions" is some esoteric combination of 7 obscure things that need to happen simultaneously.
Furthermore, I would guess that the effect of this would be similar to when one encounters bad DRAM...a system lockup or other noticeable problem. I don't think you'd just quietly get a wrong result and go on your merry way.
This is just a guess on my part, I have no visibility into this issue.
Perhaps an email into Intel support from their website asking for more specific information on this issue might yield some results.
Elmer, beamer has it right. It's a shame that the Netburst legacy will be defined by Prescott, because things really were humming along under Northwood.
Prescott was a disaster. More than anybody outside of that project will ever know. Maybe in another decade I'll share some stories. Did anybody see that disaster coming? Well, yes and no. Anybody working in the trenches knew that the schedule was a joke (a brand new design on a brand new process using brand new aggressive LVS circuit topologies and brand new cell-based design methodology, all on a compaction schedule!), and the result was that any effort on reducing power was ZBB'd to get the damn thing taped out.
What came as a surprise was how much the part's frequency was limited by power. Modelling power before you have silicon wasn't much more than a guessing game at that point. The architectural improvements actually did their job, as the IPC was comparable to Northwood despite a deeper pipeline. Performance was supposed to come from higher frequency, but no extra frequency was gained because of the power limitations (not to mention the lack of schedule to push speedpaths as hard as they could).
If more sane decisions had been made about Prescott's scope and schedule back during its planning and inception in the 2000/2001 timeframe, the general opinion about Netburst might very well be much different today. While I don't blame Barrett for Prescott specifically (I reserve that blame for the division VP at the time, who was thankfully "retired" during the recent manager cut), I do blame him for taking his eye off the ball, and taking for granted all of the risks that must be assessed before approving a project as complex as a bleeding-edge processor design.
The good news is that things feel like they're back to normal, and the general mood is more like it was back in the Wmt/Nwd days. The project goals and objectives certainly feel more sane than they ever did on Prescott, which is a very good thing.
I have a friend who's been at Intel for ~24 years now in one of the research divisions, and for a while reported to a member of the senior staff. I once asked him over lunch how he felt about Barret as CEO, and he flat-out told me that Barret's and Otellini's roles needed to be reversed (Otellini was COO at the time). He said that Barret is a great operations guy, and always has been, whereas Otellini had more vision and fresh ideas for the industry. Personally, I'm much more comfortable with Intel's current leadership, both at the corporate and VP level.
Now that April options have started trading, I've got my finger on the AMD put trigger ready to pull. I don't want anything that expires before January, and the Jan08 options are overkill, so I'm glad to see the April options available.
IMO January is the key month, and the Q4 earnings results should fully reflect the rapid 6-month Conroe ramp (negatively, of course). I think there might still be some market momentum reflected in the Q3 results.
If the price doesn't get above $26 before Q4, I'll probably wait until after the earnings announcement, and see if I can get some July puts. But if there's another positive announcement of some kind that pushes the stock even higher, I'll probably go for it sooner rather than later...
As pointed out on SI...
TigerDirect is showing 6 of their top 10 sellers as being Intel processors: 5 Core 2 Duos, and 1 Pentium D (this is up from 1 only a week ago when the Core 2 EE was the only processor readily available). It looks like they distinguish between boxed processors and mobo combos as separate items. As expected, folks appear to be flocking to the E6600...
http://www.tigerdirect.com/applications/Category/category_cpu.asp
Ya know, I've regretted not picking up some AMD puts when the first Dell deal was announced and AMD stock shot up above 40. I actually find myself looking forward to another Dell announcement, so we can have another runup based on misplaced optimism...
From that review:
"We put the AMD Athlon 64 FX-62 up against the Intel Core 2 Extreme X6800 processor. With careful optimization, the performance of an Intel Core 2 Extreme system can be increased by a substantial average of 16.8 percent (though this is where a look at our test applications is required). The Athlon 64 FX-62 only managed a performance jump of less than half that figure, 7.2 percent, with equally careful optimization. When compared head-to-head, the overclocked Core 2 Extreme outperformed its AMD counterpart by nearly 30 percent across the board."
These thrashings are almost getting boring....
I assume he's referring to WoW's fast-travel system, which involves riding a gryffon or giant bird across the country from one town to another. It's a gorgeous aspect to the game, as you fly on rails in real-time over the actual landscape of the game, but it can take a while if you're going to a distant town. Personally, when I played I usually used this time to go grab a bite to eat, but I guess I'm not the power-user Tenchu is. ;)
I sure miss WoW sometimes. Had to cut myself off a year ago because it was just getting too addictive and affecting other aspects of my life (I steer clear of MMORPGS altogether now). Great game, though...
I have to guess that they are paving the way for "Core 2 Quad" and the like. Of course, that falls apart after 4 cores. "Core 2 Octo" doesn't exactly embrace naming elegance. Say that really fast and it sounds like Klingon...