Re: My current employer just bought half a dozen single socket
NGMA based systems for circuit simulation. When we run
corner simulations we often spin off 32 or more simultaneous
Eldo or Hspice runs across the server farm. In my experience
over the last decade circuit simulation performance tracks
SPECfp with a high degree of correlation across various x86
and RISC platforms. Corner sim runs are exactly analogous
to SPECfp_rate - multiple copies of FP intensive programs
running in parallel and competing for memory bandwidth.
Well, finally! An actual real world workload example. I have been asking for one of these since the beginning of the debate, and since no one has stepped forward until you, I had begun to think they didn't exist.
The thing I would like to know is how performance tracks between Clovertown and Opteron in your particular application. Has your employer looked at Opteron as an alternative, and have they done any comparisons of performance?
I'm happy to examine the facts, but I also know that the majority of server installations are running web, database, Java, or ERP types of workloads, and there are plenty of benchmarks showing Clovertown's prowess in these segments. If there is a particular weakness in Design Simulation or Synthesis, let's investigate that separately and put it in the appropriate context.
I have no intention of letting some clueless bloke claim that SPECfp_rate is the only benchmark that matters, when such statements as these are just plain wrong. Since you know of at least one case where it is important, I'm happy to discuss that, as long as you're happy to also discuss the size of the market and the implications to Intel's overall market share in this area.
Market Data 
Markets 
AI
