Intel has won Google back as a server customer:
http://news.com.com/2100-1014_3-6153431.html
There are two things in here:
1. Intel continues to strike at AMD where it hurts them the most. On their high margin server business. As I've said below, this will restrict AMD's profits/margins and hence cash which they need to fund their ATI acquisition, their new fabs apart from all the R&D. With 16 thousand mouths to feed compared to 10 thousand a year ago on a lower revenue base and profits, AMD has to cut back.
2. Intel is now displaying a remarkable flexibility to customize solutions for individual customers. I had mentioned in an earlier post about how the microprocessor and car industry business models would converge in the future. This is one part of what I was referring to. More about it in a later post.
Till then...AMD won TCL in PRC. But probably to fight Lenovo on price which means low margins for AMD. Meanwhile, Intel continues to take high margin server business away from them with the Google/Sun wins. I need to scrutinise the AMD numbers a little more closely before I can commit in a post but I wouldn't be surprised if they head to an operating loss in Q2/Q3 of this year.
Subscribe to:
Post Comments (Atom)
7 comments:
While some of these server announcements are nice PR - they are only good for Intel if it represents an overall trend. The Google/Sun business in and of itself is relative peanuts, but it does point out people's willingness to change quickly in server space these days (which is a bit surprising to me as this has traditionally been a rather conservative area).
I did find it rather interesting that AMD said they don't expect K8l to hit the books significantly until Q4 - this means paper launch in Q2, ramp volume in Q3 and get significant volume in Q4. This is very similar to Core2 which went from launch in June to ~75% of Intel's server production by end of Q4'06.
So the question is how much server share will Intel be able to eat back in the server space over the next 3 quarters and how will AMD have to price K8l in an attempt to win back business. I suspect the old opteron pricing days are now gone...
In my view mobile is what will or will not keep AMD in the green (pun intended) over the next 3 quarters as I think Intel will continue to cut into server share and/or force further reduction in AMD ASP's to keep the server market share. Either way this will hit gross margin and the only thing that I think could possibly offset this is sustained mobile growth.
I think the other thing to consider about your operating loss prediction is whether or not stock compensation expenses will be factored in or continued to be split out and counted in the GAAP calculations.
http://www.channelregister.co.uk/2007/01/28/google_amd_xeon/print.html
And yet...
while somebody having ballons and parties, somebody will just came into the stage and said disturb that...
Nice going, Intel, but you've to try harder, and Sharikou180, sorry to interrupt your high spirit for Intel.
Did you read the article I linked to on Cnet. It already has the quote from Google you've linked to.
Can't see anything for Google's response on that page though, just ballons and celebrations of fanboys. Too bad.
P.S. maybe it's my short-sighted eye to have messed up my internet browsing experience.
Sadly, Google felt that it needed to state that it only bought a small number of Intel chips
I doubt that Google was buying Intel machines for production server use. My best guess is they were buying them as test systems. Google must do extensive QA for their new products. As Intel has the majority of the desktop market, Google must test their new products to make sure they can run properly on under performing Intel desktops.
Post a Comment