0.7 M cEMs/sec for Pentium M...

[Gruznov]

I'm perhaps stupid but I cannot understand why sob programme (Win 2.5.0) produce 0.7 M cEMs/sec and nothing more (Pentium M Celeron 1.4 Ghz). PentiumIII count faster!!! I'm inclined to believe that programme code is a thorough sh*t and not Intel's CPU. Or alternatively I'm single and no one have such problem...

:swear:

[Matt]

If you're running secondpass your cEM/s will appear lower.

[pixl97]

Is your PM in a laptop? are you sure its running at full speed and not in low power mode, they can kick down to just a few hundered MHz in power saving mode, get CPU-Z and see what its running at. And also, are you sure there are no other programs sucking up processor time.

[Ken_g6[TA]]

I have a Celeron M 1.5GHz I just got recently, that I'm running on very small numbers at the moment. It seems to be about 15% slower than my Athlon XP 1.8GHz. (I'm ignoring the CEM/s, and just looking at the time per block, for comparable size numbers.)

I was wondering, though, since George Woltman's software has had problems with the Pentium M and SSE2 in the past, whether (1) SB uses SSE2 on Pentium M, and (2) whether anyone has tried comparing the SSE2 code with the generic code on Pentium M?