SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Advanced Micro Devices - Moderated (AMD)
AMD 248.09+6.2%10:34 AM EST

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: david1951 who wrote (244549)12/17/2007 12:06:45 PM
From: Sarmad Y. HermizRead Replies (1) of 275872
 
>> Have read conflicting reports about the patch affecting performance. Not a tech guy but who is doing the testing seems to have a lot to do with the results.
<<

David, look specifically for tests on virtualization work loads. These are more important than any other. Without that aspect fixed, the chip will not be sold (for servers), period.

siliconinvestor.com

Posted by: chipguy
In reply to: smooth2o who wrote msg# 54815 Date:12/17/2007 10:37:40 AM
Post #of 54826

Did you notice this?

Ashlee wrote:

Speaking of performance issues, we've learned that virtualization software suffers from AMD's fix. The Register has received reports of virtualization software slowing by as much as 50 per cent.

In the commentary some one claiming to be from AMD wrote:

I head up AMD's XEN testing, and direct comparison testing (same changeset, same bittedness, same guests) against same speed XEON systems aren't indicating any sort of speed difference, let alone 50%!

Curtis W. Rendon

AMD, Austin

Do you notice the difference? Ashlee claimed that K8L
took a 50% hit from the TLB bug fix on virtualization. This
guy says that clock for clock K8L is about as fast as Xeon
for virtualization. Aside from Ruiz class obfuscation and
misdirection in appearing to refute Ashlee's claim but not
really doing so, wasn't AMD claiming a huge performance
advantage over Xeon in virtualization for K8L about a year
ago? Now it's not there even clock for clock.
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext