SiSoft Sandra Professional 2005

Dhrystone ALU (higher is better)

No surprises here. When it comes to integer performance, the Opteron’s are incredible. It’s amazing that Intel’s Dual-Core Pentium EE 840, a desktop processor, is about on par with our “ultimate” workstation setup from about 2 years ago.

Whetstone FPU (higher is better)

Intel CPUs always do well here, although it’s odd that real-world experiences always seem to put AMD ahead in the FPU arena. Still, for customized non-SIMD software the Pentium lineup is a strong contender.

Integer SSE2 (higher is better)

Floating Point SSE2 (higher is better)

Most of today’s advanced scientific computing platforms use SSE for calculations. Intel has traditionally held the advantage when it comes to SSE processing, but the newer Opteron steppings provide a strong showing. Again, it is impressive to see how well Intel does, but the flagship performance is with the 2x Dual-Core Opteron 275’s.

Matlab N72 – Time to completion (shorter is better)

It seems like Mathwork’s original claim that most Matlab scripts are limited by the parsing of the script is true. Here, the added clockspeed of the Opteron 252’s push it to the very top. Though the script includes a moderate amount of matrix math, it doesn’t seem like much of it is parallelized. Our recommendation from two years ago still stands – for most Matlab users, the fastest performance will come with a single Athlon64 line. We hope to follow-up in the future with additional Matlab tests once AMD’s own Math Core Libraries are available. We did not have time to evaluate the Intel platform with the Intel MKL, the P4 3.0GHz is an older reference measurement.