Hey there.
I'm sure this has been mentioned, but I just noticed that the Big Mac cluster up to 3rd place, moving past the Itanium2 cluster (and it's even edging the Itanium cluster on GFlop/s per CPU: 4.52 to 4.45).
I'd say this is as high as it will go, seeing as #2 is a Linux cluster with 8,160 CPUs that's churning oput almost 14,000 GFlop/s.
The G5 cluster is getting close to 10,000 GFlop/s, with 9,555.
I think you can draw the conclusion from this that both the PPC 970 and Itanium2 are pretty kick-ass chips, and it's nice to see IBM/Apple have the edge for once!
The PDF file is here.
Edit: I forgot to attribute Macs Only! for the story.
I'm sure this has been mentioned, but I just noticed that the Big Mac cluster up to 3rd place, moving past the Itanium2 cluster (and it's even edging the Itanium cluster on GFlop/s per CPU: 4.52 to 4.45).
I'd say this is as high as it will go, seeing as #2 is a Linux cluster with 8,160 CPUs that's churning oput almost 14,000 GFlop/s.
The G5 cluster is getting close to 10,000 GFlop/s, with 9,555.
I think you can draw the conclusion from this that both the PPC 970 and Itanium2 are pretty kick-ass chips, and it's nice to see IBM/Apple have the edge for once!
The PDF file is here.
Edit: I forgot to attribute Macs Only! for the story.