processor dependent math results

[George_P_Burdell]

Oskar-

I’m glad we were able to get everything sorted out for you.  I will be creating a KnowledgeBase Article concerning this for future customers who enquire about this.  Good luck with the rest of your application.

[DSPmchen]

Hi Oskar,

Glad to see the solution works for you!

Since the solution will introduce additional data copy of sub-matrix, I am not surprised that the solution is 1.6x slower than the build-in A x B VI on Core 2 machine.

MKL uses different optimized code on different processors.  I know Intel designs specific code for Core 2 CPU.  So it is possible that the VI  on Core 2 is faster than Xeon Processor.

Best Regards,

Michael