Hawkeye_(BEL)
Banned
Hi all,
I seem to have a problem with S@H. I think it's running too slow, but I'm not sure, since I've heard they were handing out some 'difficult' WU lately.
Anyway, here are the specs of the PC :
- PIII800E@880
- Via Apollo Pro A mobo
- 128Mb PC133 Cas3 @ 110 Cas2
I'm running the winnt-cmdline 3.0 version with setispy (latest version). Now, I get WU times ranging from 7 hours to 5.45 hours, on 6 different WU's tried. I remember that the 2.x clients liked a big, fast L2 cache and a high memory bandwidth, but I don't know what the 3.x clients like.
So I tried to tweak the chipset a little (as Via chipsets are known to have poor memory performance) with WPCREDIT, and I got an increase of 100 points in the Sandra memory benchmark, but the S@H times were unchanged.
Now, setispy is reporting I've a CpF (whatever that may be) of 6.76, and that a normal PIII Coppermine should get a CpF of 5.1. Now, a friend with a CeleronII 566 does a S@H WU in the same time as I do. So there has got to be something wrong.
I don't run any virus scanners, I checked with wintop for any possible programs thay may be eating up CPU power, but I've found nothing of significance.
Can someone give me some hints on what I can try to improve performance ?
Thank you.
PS : is it better to run the CPU at 800 and the memory at 133 again for S@H ?
I seem to have a problem with S@H. I think it's running too slow, but I'm not sure, since I've heard they were handing out some 'difficult' WU lately.
Anyway, here are the specs of the PC :
- PIII800E@880
- Via Apollo Pro A mobo
- 128Mb PC133 Cas3 @ 110 Cas2
I'm running the winnt-cmdline 3.0 version with setispy (latest version). Now, I get WU times ranging from 7 hours to 5.45 hours, on 6 different WU's tried. I remember that the 2.x clients liked a big, fast L2 cache and a high memory bandwidth, but I don't know what the 3.x clients like.
So I tried to tweak the chipset a little (as Via chipsets are known to have poor memory performance) with WPCREDIT, and I got an increase of 100 points in the Sandra memory benchmark, but the S@H times were unchanged.
Now, setispy is reporting I've a CpF (whatever that may be) of 6.76, and that a normal PIII Coppermine should get a CpF of 5.1. Now, a friend with a CeleronII 566 does a S@H WU in the same time as I do. So there has got to be something wrong.
I don't run any virus scanners, I checked with wintop for any possible programs thay may be eating up CPU power, but I've found nothing of significance.
Can someone give me some hints on what I can try to improve performance ?
Thank you.
PS : is it better to run the CPU at 800 and the memory at 133 again for S@H ?