Read these two articles on Uberpulse (thanks Anonymous for the link):
8 Intel cores faster than 16 AMD cores
Barcelona recent lack of performance and why AMD should start to worry
It is extremely concerning that AMD continues to be unspecific about the Barcelona launch date (I mean come on...we're a few days short of mid 2007) and the absence of samples for the tech press to do independent benchmarking persists.
Subscribe to:
Post Comments (Atom)
5 comments:
Ive never said something like that, but I think now the time has come:
AMD BK not very far away.
I see you jumped on the same wrong way bandwagon. The AMD demo doesn't use SSE for Pov-Ray. With SSE, K10 is much faster. At the same clock and compiled properly for AMD, it is faster than Penryn.
Scentia doin' damage control again? More BS from Scentia again.
Lets say it isn't using SSE... Didn't AMD said K10 is 50% faster in FP only? So, if K10 is only using FP then why is it not faster than K8? You can deduce this from the demo "running at the same frequency".
And don't give us that TDP crap. If it was a TDP thingy, where is power or watt meter at the wall? AMD didn't show that. Allen was busy talking about scaling most of the time!
Scaling demo or not, it shows us K10 isn't that much faster than a K8. And doesn't scale perfectly either. 82% from 4 to 8. Let's not forget about 86% at 1 to 4 cores. In other words at 1 to 8 cores, its only 68% scaling.
Henri's excuse was a joke.
http://theinquirer.net/default.aspx?article=39896
V8 gets a higher score than K10 16-cores? Oh man this is very funny.
Scientia is far too late to do damage control.
If anon or this blog's author actually thinks a comparable Intel chip does anything 2x faster than its AMD counterpart, you guys are too far gone to help and are sufficiently damaged by PR. You guys are probably scratching your heads wondering how Intel, with its twice faster tech. hasn't already bankrupted sorryful AMD. Use some common sense.
Post a Comment