AMD losing lucrative workstation share to Xeon:
Link
This number is going further south this quarter. AMD will probably end back up at 6 points or so. I already said it - they will still lose money this qtr. Probably around half what they lost in their disastrous Q107.
They need to intro Barcelona fast...but more importantly their biggest medium term hope is Fusion. Until then, Intel will continue to have them on the ropes.
Subscribe to:
Post Comments (Atom)
8 comments:
"..but more importantly their biggest medium term hope is Fusion"
Really? Isn't fusion more of a low end market product initially? Do you think people buying high end workstations want integrated GPu capability that will not likely be as good as a discrete solution, and also not upgradeble without throwing the chip away too?
As a higher end desktop use, all things being equal I would rather have a discrete graphics solution that I can choose to upgrade at a different pace / timing then my CPU. I also would wonder if having the GPU on board would mess with OC potential for the CPU?
Of ocourse if fuwion offered BETTER performance than discrete solutions my mind might change, but simply equivalent performance (which I think is the best AMD can hope for initially) would not outway some of the limitations of this approach (flexibility). This is also why I think it is more targeted at mid-low range products or mobile where people do not intend to upgrade inidividual parts as much, or put as much importance on having a high end grpahics solution.
I have no direct evidence that the integrated GPU will not be better than discrete high end solutions, but I think that is a a reasonable assumption (as least on initial product).
more specifically, AMD lost it because it did not have Quad-Core.
The song and dance about "native" quadcore not only did them no good in terms of loss of sales but also massive production delays and wasted material due to forcing a monolithic solution on a now obviously immature process node.
Big "i told you so" from intel directed at AMD for this major strategic blunder.
My Fusion comment wasn't in the context of workstation. It was more around it allows them to further segment the market as the measurement of performance will become more ambiguous when you add graphics to the silicon.
As a result, AMD will be able to segment the market further and Intel cannot chase them everywhere...it takes too much resource to do that.
180:
Just a quick question. Dell inked a deal w/ Wal Mart to put Dells in retail @ WM. The bulk of those SKU's as I understand it will have AMD as the processor.
My quick take on this is that AMD is getting share where they need it least, the value PC market.
3 reasons or so why this is short-term good, long-term bad:
1) Eating precious capacity that is not returning much or any margin.
2) Customer base that is less well-informed getting a lower quality PC and blaming AMD; thus Intel on their next purchase, possibly.
3) Dell warranty issues; will customers be confused about their warranty/tech support and will it be handled differently than a direct-sales approach?
Thoughts?
"The song and dance about "native" quadcore not only did them no good in terms of loss of sales...."
It's also going to come around and bite them in the a$$ from a PR perspective, if they do plan to go the MCM route in the future with 2 native quad cores.... I wonder what the marketing spin will be then...
(Although I'm sure the fanboys will be out in full force talking about
brilliant this strategy and implementation is)
"Our die are glued more tightly"
"We use better glue"
"Please refer to out octo-core for dummies manual"
"Intel copies us all the time, we though turn about was fairplay"
"We use a 'true' MCM design"
"Our process and yield is still more advanced, we just chose not to take advantage of it by using a native design..."
On the Walmart/Dell issue, I had a look at the systems online at Walmart. Dell is offering 2 SKUs which are basically a low end and a slightly more bundled version of the same system.
Features the A64 X2 3600+. The lower end system has no monitor/speakers, smaller hard disk and costs $498.
The more beefy system offers a 19" flat panel, speakers, larger hard disk at $698.
If you go to dell.com and try and configure the same system you get:
1. Spec to spec the $698 system is $769 on the Dell site.
2. The $498 system is $599 on the Dell site. However, you gotta take a 15" flat panel at least so the delta is not that large for this system.
However, Dell is definitely discounting these to Walmart.
My take is AMD doesn't have a choice...or any say. Dell would have a contract that allows them superb pricing but they would also have to buy certain quantities. Hence, the fact that Dell is only offering 2 SKUs of the same basic system featuring the 3600+ tells me Dell is using this deal for now to move static inventory. Dell actually has systems that start at $359 so if the bottom is $498 at Walmart then it's unlikely they're trying to win back market share. They're probably testing the waters and at the same time trying to move inventory that will have zero value if they can't get rid of it very soon.
The cheapest 3600+ on Newegg is retailing at $59. So you can imagine what price Dell would have scored these. AMD is probably losing money on these parts. But it's better to sell a part at a loss than write it off. I don't think capacity is the issue because AMD has too much capacity (like Intel does too). Also, these are probably parts already committed to Dell. The upside for them is brand presence at a new point of sale riding under Dell's brand cred.
If there is down-side for anyone, it's going to be Dell. If consumers find out that Dell is offering the same systems cheaper through a retailer, it defeats the entire logic of their direct model which is passing the savings back to consumers by cutting out the middle-man.
Post a Comment