[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Some Linux performance
On Tue, 14 Apr 1998, Tuvik Beker wrote:
> - 2 333MHz P-II CPUs
> Running the new "PC", the same simulation finished in 5:40 minutes. The
> amazing fact is that running two identical processes simultanuously
> resulted in exactly the same time - for BOTH, that is. That means that
> (in my case at least, under rather sterile conditions) the linux SMP
> code gave an amazing full 100% speed increase.
that's cool, as well as pretty close to impossible. I can only suppose
your demo was pretty intensive on the user space and not so heavy on
kernel space. plus your data was possibly all in cache and no massive disk
WRITING was done, plus you haven't mentioned if you have the old klunky
2.0.3x kernel or one of the latest 2.1.9x kernels with better SMP support.
For the record, I'd like to mention here I am also jelous, but not too
much. I have a dual P6 233 here and it functions faster than any other
machine here too (SGI, IBM, and a dual Ultrasparc 266)