FUDzilla claims Barcelona is delayed till Q4. (I am unclear on what basis).
http://www.fudzilla.com/index.php?option=com_content&task=view&id=1351&Itemid=1
http://www.fudzilla.com/index.php?option=com_content&task=view&id=1351&Itemid=1
The AMD Execution Thread [2007 - 2017]
- Thread starter overclocked_enthusiasm
- Start date
-
- Tags
- amd
- Status
overclocked_enthusiasm
Regular
I saw that also. There certainly is alot of different sources all talking delay with a lone low level AMD spokesman saying everything is fine. If you were AMD wouldn't you try and squash these rumors if they were untrue? In fact, wouldn't you have shown a higher clocked part instead of 1.6 GHz to prevent these rumors from taking root in the first place? Something doesn't add up and we have all seen what that generally leads to.
Bouncing Zabaglione Bros.
Legend
There was a financial report from Cray last week that let the cat out of the bag. IIRC, Cray won't be able to sell their Barcelona based machines because according to them, Barcelona has slipped back to the end of the year.
Barcelona or Budapest?
I thought Cray's delay was for products that used HT3, which Barcelona does not have.
I thought Cray's delay was for products that used HT3, which Barcelona does not have.
Bouncing Zabaglione Bros.
Legend
You're right, I misremembered. It's Budapest that Cray are waiting on, but AMD are still insisting that Barcelona is on schedule. Let's hope that's not a R600 style "on schedule".
That might depend on the set of rumors are correct.
Some rumors have it that the demoed Barcelona chips are buggy, perhaps more buggy than a chip this close to release should be.
New steppings would take time, though. If a new stepping is progressing through the fab right now, it would be weeks before it would be known if the bugs were fixed, then weeks before volume production.
If these were old stepping Barcelonas, a new stepping would be further along. Some are saying that these are recent stepping chips, however.
Some rumors have it that the demoed Barcelona chips are buggy, perhaps more buggy than a chip this close to release should be.
New steppings would take time, though. If a new stepping is progressing through the fab right now, it would be weeks before it would be known if the bugs were fixed, then weeks before volume production.
If these were old stepping Barcelonas, a new stepping would be further along. Some are saying that these are recent stepping chips, however.
http://www.fudzilla.com/index.php?option=com_content&task=view&id=1360&Itemid=1
ATI to get 60 percent of OEM deals
ATI to get 60 percent of OEM deals
overclocked_enthusiasm
Regular
http://www.gearlog.com/2007/06/amds_barcelona_16ghz_and_disap.php
"Back to Barcelona: DailyTech reported that partners were talking about speeds of 2.0-GHz. I didn't get that (I heard a "possibly 1.8-GHz at launch"), but we are still (unfortunately) talking about early silicon. What's known as the "B0" stepping is out, and the "B02" stepping should be available shortly, I was told."
"Back to Barcelona: DailyTech reported that partners were talking about speeds of 2.0-GHz. I didn't get that (I heard a "possibly 1.8-GHz at launch"), but we are still (unfortunately) talking about early silicon. What's known as the "B0" stepping is out, and the "B02" stepping should be available shortly, I was told."
I think it's worth noting that really, AMD is very competitive with the A64 in desktops.
I know, I'm looking to rebuild my PC sometime in the next few months, and AMD has strong, often superior contenders to the core2dou at any price point. This is something most people seem to be ignoring while beating the "Intel is killing AMD on performance right now!" drum. It's simply not true. I'd rather go with AMD, also, because they have cheaper more stable third party boards that I'm long familiar with.
The kicker though is the immense overclockability of the C2D's. This will be most likely why I go with C2D, the ability to buy a $130 chip and OC it to near top of the line, that AMD just doesn't give you.
HOWEVER, it remains that the vast, vast majority of laypeople will never overclock, so it's not a factor to the vast majority of consumers nor ever will be. To them AMD is very competitive right now.
I was just poking around newegg for example, and right now the X2 5400+ (2.8 ghz) is listed at $155. Without overclocking, Intel does not have a chip that can compete at that price point with that performance. E6400 might get close, but at ~$30 more.
I know, I'm looking to rebuild my PC sometime in the next few months, and AMD has strong, often superior contenders to the core2dou at any price point. This is something most people seem to be ignoring while beating the "Intel is killing AMD on performance right now!" drum. It's simply not true. I'd rather go with AMD, also, because they have cheaper more stable third party boards that I'm long familiar with.
The kicker though is the immense overclockability of the C2D's. This will be most likely why I go with C2D, the ability to buy a $130 chip and OC it to near top of the line, that AMD just doesn't give you.
HOWEVER, it remains that the vast, vast majority of laypeople will never overclock, so it's not a factor to the vast majority of consumers nor ever will be. To them AMD is very competitive right now.
I was just poking around newegg for example, and right now the X2 5400+ (2.8 ghz) is listed at $155. Without overclocking, Intel does not have a chip that can compete at that price point with that performance. E6400 might get close, but at ~$30 more.
Last edited by a moderator:
AMD is price competitive in the mid to low end of the market it is capable of supplying, but not cost-competitive.
AMD's 90nm top-bin chips compete well on price/performance with Intel's smaller, cooler, more cheaply produced and higher margin mid-range processors.
That's good for consumers, but AMD's shrunken margins are testament to how undesirable the situation is from a business standpoint.
AMD's 90nm top-bin chips compete well on price/performance with Intel's smaller, cooler, more cheaply produced and higher margin mid-range processors.
That's good for consumers, but AMD's shrunken margins are testament to how undesirable the situation is from a business standpoint.
The most disturbing aspect of AMD's current financial condition is the issue of pricing. As we know, AMD resorted to price-slashing in order to combat the performance advantage Intel had (to whatever degree). Since the arrival of A64, gaming has traditionally favored AMD by a large margin and that margin suddenly flipped with the arrival of Intel's Core2Duo. So, with prices being slashed....margins were also slashed....meaning profits slashed......despite driving solid volume.
In short, AMD desperately needs to return to appreciable margins...much higher than what they're seeing currently. Opteron used to be a "magic-bullet" here thanks to the high prices of enterprise hardware and extremely high margins. However, with Intel backing off their Itanium push and focusing on delivering solid items such as 45u Harpertown Quad-Core in Q4 2007, AMD is forced to look elsewhere for help with recovering margins...Granted, they are hoping for native quad-core to provide relief...but yields and production costs are certainly worse-off than they'd hope for (if rumors are correct).
Clearly, AMD's true hope is to have a commanding performance lead which will justify charging more for their processors. Unfortunately, Intel has more 45u processors waiting in the wings and Penryn looks like a very solid challenger to AMD's Phenom processors. Early indications seem to favor Intel and rumors of a delay for Barcelona definitely don't aid AMD's situation...
AMD wants, needs, and hopes for another "Opteron" or "Athlon64" product which will catch the industry by surprise and blow away expectations...Consumers should be hoping for the same because competition is what helps keeps pricing reasonable and continues to drive innovation. If AMD drops out of the picture (doubtful, but dangerously close it seems these days), the industry could be in some serious trouble...
In short, AMD desperately needs to return to appreciable margins...much higher than what they're seeing currently. Opteron used to be a "magic-bullet" here thanks to the high prices of enterprise hardware and extremely high margins. However, with Intel backing off their Itanium push and focusing on delivering solid items such as 45u Harpertown Quad-Core in Q4 2007, AMD is forced to look elsewhere for help with recovering margins...Granted, they are hoping for native quad-core to provide relief...but yields and production costs are certainly worse-off than they'd hope for (if rumors are correct).
Clearly, AMD's true hope is to have a commanding performance lead which will justify charging more for their processors. Unfortunately, Intel has more 45u processors waiting in the wings and Penryn looks like a very solid challenger to AMD's Phenom processors. Early indications seem to favor Intel and rumors of a delay for Barcelona definitely don't aid AMD's situation...
AMD wants, needs, and hopes for another "Opteron" or "Athlon64" product which will catch the industry by surprise and blow away expectations...Consumers should be hoping for the same because competition is what helps keeps pricing reasonable and continues to drive innovation. If AMD drops out of the picture (doubtful, but dangerously close it seems these days), the industry could be in some serious trouble...
AMD's execs and engineers aren't stupid, though.
By this point, there is no reason for AMD to just hope for another Opteron, because they know very well how Barcelona performs.
Their silence is not encouraging, and they've played very fast and loose with the numbers.
They wouldn't be doing this if things were going well.
By this point, there is no reason for AMD to just hope for another Opteron, because they know very well how Barcelona performs.
Their silence is not encouraging, and they've played very fast and loose with the numbers.
They wouldn't be doing this if things were going well.
overclocked_enthusiasm
Regular
Now that Intel has gotten away from Netburst, their superior fab and manufacturing might is beginning to create distance betwen themselves and AMD. In addition to the advantages in maufacturing that sheer scale create for Intel, their much bigger R&D budget is beginning to bear fruit as well. The real reason for AMD's success the past several years was as much about Intel shooting themselves in the head with Netburst as AMD bringing out Opteron and A64.
It reminds me a little of ATI vs Nvidia. One could argue that R3xx was the reason for ATI's dominance and others would argue R3xx shined so bright because NV30 sucked so bad. Since the NV30 vs R3xx period Nvidia has shown that they are a much more effecient company in terms of product launches, gross margins, profitability and marketing. So the appearance is (rightly or wrongly) that ATI just got lucky and R3xx was a one hit wonder since they have been unable to duplicate the success of it from a technological or financial viewpoint since.
The same arguement is being made now with AMD. It took AMD almost 2 years to claw their marketshare up to 25% compared to Intel by the end of Q4 2006. This was due to the success of Opteron and A64 to be sure. But it was also due to the fact that Intel was stuck with Netburst before Conroe showed them the way out of their misery. In just 90 days with Core2Duo Intel dropped AMD's marketshare from 25% back down to 19% in Q1 2007. 2 years of hard work evaporated that fast for AMD.
Yes their was an oversupply of the OEM s and a sense of abandonment for the channel which contributed to this huge marketshare swing. Remember when ATI did the same thing by going full PCIe way before the industry was ready? Not only did they completely overbuild their inventory on PCIe parts before they were needed but they left the AGP dominated channel out to dry. ATI never fully recovered from that financially and they ultimately had to writeoff much a large portion of inventory as a result.
AMD's inventory swelled by about 20% in the first 90 days of 2007 to over $900 million. 65% of this inventory is those 90 nm parts that are likely showing up in the recent Wal-mart announcements via Dell. While this will certainly help AMD move excess inventory the pressure on gross margins, debt, cash flow and other financial metrics will continue.
The only way out of the hole will be as other posters have stated...regaining the performance/technology lead and the pricing power that is associated with it. Anything short of that will be a slow bleed of limited cash resources that AMD possesses and the constant pressure from Intel on die shrinks and manufacturing effeciencies will leave AMD with very limited options.
In the end I think the most likely outcome is a PE buyout or AMD spinning off assets to raise cash. I don't see how an asset light approach will allow AMD to combat Intel's fab might and I doubt TSMC or Charter will be able to produce the chips AMD needs to compete in time or with gross margins high enough to make the current AMD business model viable. I will say it again...Barcelona or bust.
It reminds me a little of ATI vs Nvidia. One could argue that R3xx was the reason for ATI's dominance and others would argue R3xx shined so bright because NV30 sucked so bad. Since the NV30 vs R3xx period Nvidia has shown that they are a much more effecient company in terms of product launches, gross margins, profitability and marketing. So the appearance is (rightly or wrongly) that ATI just got lucky and R3xx was a one hit wonder since they have been unable to duplicate the success of it from a technological or financial viewpoint since.
The same arguement is being made now with AMD. It took AMD almost 2 years to claw their marketshare up to 25% compared to Intel by the end of Q4 2006. This was due to the success of Opteron and A64 to be sure. But it was also due to the fact that Intel was stuck with Netburst before Conroe showed them the way out of their misery. In just 90 days with Core2Duo Intel dropped AMD's marketshare from 25% back down to 19% in Q1 2007. 2 years of hard work evaporated that fast for AMD.
Yes their was an oversupply of the OEM s and a sense of abandonment for the channel which contributed to this huge marketshare swing. Remember when ATI did the same thing by going full PCIe way before the industry was ready? Not only did they completely overbuild their inventory on PCIe parts before they were needed but they left the AGP dominated channel out to dry. ATI never fully recovered from that financially and they ultimately had to writeoff much a large portion of inventory as a result.
AMD's inventory swelled by about 20% in the first 90 days of 2007 to over $900 million. 65% of this inventory is those 90 nm parts that are likely showing up in the recent Wal-mart announcements via Dell. While this will certainly help AMD move excess inventory the pressure on gross margins, debt, cash flow and other financial metrics will continue.
The only way out of the hole will be as other posters have stated...regaining the performance/technology lead and the pricing power that is associated with it. Anything short of that will be a slow bleed of limited cash resources that AMD possesses and the constant pressure from Intel on die shrinks and manufacturing effeciencies will leave AMD with very limited options.
In the end I think the most likely outcome is a PE buyout or AMD spinning off assets to raise cash. I don't see how an asset light approach will allow AMD to combat Intel's fab might and I doubt TSMC or Charter will be able to produce the chips AMD needs to compete in time or with gross margins high enough to make the current AMD business model viable. I will say it again...Barcelona or bust.
Eh, AMD does not need the performance lead to stay alive. They did it long ago without it. Granted, everybody didn't wish for them to fail as is the case today back then.
They'll just have to settle for lower margins than Intel. No problem at all, if they're not stupid. I'm sure Intel is a very expensive company. There comes a certain amount of that with being so large. AMD should be able to cost undercut them pretty easily, again, IF they're not stupid, which they've shown plenty of intention to be stupid..
They'll just have to settle for lower margins than Intel. No problem at all, if they're not stupid. I'm sure Intel is a very expensive company. There comes a certain amount of that with being so large. AMD should be able to cost undercut them pretty easily, again, IF they're not stupid, which they've shown plenty of intention to be stupid..
overclocked_enthusiasm
Regular
The problem with this line of thinking is that AMD has set itself up to be able to supply 30% of the CPU market with their CapEx budget, Fab buildouts and flex capacity at Charter. They have made the decision to compete with Intel head on and not be relegated to the low end scraps as you suggest. In order to fill this Fab capacity and to support their CapEx plans, they need to sell X amount of product at decent gross margins. Without the performance lead they can MAYBE do the first but certainly NOT the latter. If Q1 2007 is any indication, they can't do either without the performance lead and will never get above 20-22%marketshare. This sort of business model will lead to losses of hundreds of millions each and every quarter and bleed the company dry in short order.
The economies of scale that you mention work in favor of Intel and against AMD...not the other way around. Since AMD now only has 19% of the market instead of the 25% they had just last quarter this becomes an even bigger problem for AMD and a better and better situation for Intel. As Intel is able to maximize their Fab capacity they are able to be much more effecient and their gross margins rise...the converse is true for AMD. Remember these are not Fabless companies like ATI and Nvidia they are VERY capital intensive companies that must invest billions each year to accomodate die shrinks, new fabs and the R&D associated with chip manufacturing.
The way that AMD is currently capatilized, the Fabs they have in production and the CapEx plans they have laid out for 2007 and beyond they are not trying to be the low cost low end supplier of the masses. They are trying to regain the performance lead for the very life of the company. Without the performance lead AMD's current business model is worthless and the company would have to restructure. In addition, Intel would have to cede the lowend to AMD which they may be unwiling to do regardless of the gross margins involved.
Remember, the performance lead by itself is meaningless. However the gross margins, marketshare penetration and halo effect that surronds it are the holy grail of any tech company. Just look at Apple and the Ipod or Intel and Core2Duo as examples of what the Halo effect, marketshare and gross margin impact having "the best" has on a company. If Barcelona nd Phenom are not better than Penryn AMD is sunk and may not survive long enough to try and introduce Fusion. Each quarter that passes now will bleed AMD of its finite cash reserves. This can go on for another 4-6 quarters MAX before AMD will be BK.
The thing to remember here is that this mode of thinking applied to the pre-Opteron/Athlon 64 AMD....when AMD was the consumate underdog. The financial industry had very reasonable expectations for the company and was satisfied with a low market share thanks to solid margins.
Unfortunately for AMD...the wild success of Opteron/Athlon 64 and the associated market share gains created a "Cinderella story" effect for the stock. Reasonable or not, the financial industry now expects AMD to be able to combat Intel's best efforts on every product cycle. In the case of K10, the financial industry has been led to expect another K8-level revolution thanks to the urging of AMD's own execs. So, AMD is facing what seems to be some unrealistic performance expectations. When you add the lofty expectations AMD's acquisition of ATI created, you see AMD has an extrememly tough uphill battle just to meet expectations.
In short, AMD's situation comes down to money. If margins and market share are low, the stock will get hammered. If expectations aren't met, the stock will get hammered. If the stock gets hammered too much, AMD will be bleeding money and need to make some drastic changes. What those changes are is obviously unknown, although you can be certain AMD will be a vastly different company than the one we all grew to love as the "ultimate underdog" doing battle with the 800lb gorilla.
Interesting times for sure...
When you consider the headcount cuts Intel did last year when they were actually in the process of getting healthy, it seems to me AMD has made only token cuts so far. Doesn't mean it won't be painful for those involved, of course. But as quick as they've grown in recent years, it's hard to not think that they could cut 10% without hitting bone.
Normally i wouldn't link to these rumors, but after reading it i was left pondering the financial implications of this hypothesis from the AMD perspective:
http://www.hardspell.com/doc/hardware/38005.html
Google "sketchy" translation:
Unlikely as it seems (and i think this is extremely unlikely, mind you), how much would AMD stand to lose in the immediate term ?
How much would Nvidia stand to lose/gain ?
And what about a possible motivation for this sudden "twist" ?
http://www.hardspell.com/doc/hardware/38005.html
Google "sketchy" translation:
Unlikely as it seems (and i think this is extremely unlikely, mind you), how much would AMD stand to lose in the immediate term ?
How much would Nvidia stand to lose/gain ?
And what about a possible motivation for this sudden "twist" ?
- Status
