OpenMP is a system independent set of procedures and software that arranges automatic parallel processing of shared memory data when more than one processor is provided. This option is available in the latest Microsoft C++ compilers. The new benchmark (see Files) executes the same functions, using the same data sizes, as the CUDA Parallel Computing Benchmark, with varieties compiled for 32 bit and 64 bit operation, using old style i387 floating point instructions and more recent SSE code. A run time Affinity option is available to execute the benchmark on a selected single processor.

Can anybody see what happens with a quad (or more) cpu and attach the text log file results to a msg?

Roy

Attachments
	Name:   Results_V64.txt	Size:   3 K

Attached are 32 bit and 64 bit compilation results of my OpenMP benchmark on a Phenom II X4 CPU using 64-Bit Windows 7. The 32 bit version uses old i387 floating point instructions with the 64 bit variety running SSE instructions, but not very efficiently due to compiler limitations.

The main observation is that, with 4 CPUs, a single program using OpenMP can produce speeds almost four times faster than a single CPU, where there are not cache size limitations.

Note software information shown by the benchmark:

Windows 7 is identified as Version 6.1, build 7600.

64 Bit Version correctly shows 8 GB RAM and 8388 GB Virtual Space

32 Bit Version at least shows 4 GB RAM available and 4 GB Virtual Space

Roy

Attachments
	Name:   OpenMP32.txt	Size:   3 K
	Name:   OpenMP64.txt	Size:   3 K