SOB: ....

[Confused]

Glad you got it figured out


Confused

 

[InverseOfNeo]

They seem to be going slower than when I tried it without the service. I was getting about 120k cems/sec when it was using 50% of each cpu (obviously the config wasnt correct to fully utilize both cpus), Now I am getting about 70k cems/sec each. On a different note my P4 rig seems to get just under 295k cems/sec

Does that seem right or is something still wrong?

 

[MathGuy]

Hi, InverseOfNeo - if you send me the logs (both sobsvc.log and sb.log) as well as some general details on your system (OS, # and type of processors), I'll try to help you troubleshoot it. My e-mail is in the readme that came with the service handler, or you can reply via this forum.

 

[InverseOfNeo]

Thanks MathGuy, I think I got it. I didnt specify the # of cpus for the service (i.e. sobsvc -p:2). I did that and restarted the service and it seems to be going well now. If I have trouble with it later on then I will get your help with it. Thanks again.

 

[Baldy18]

I let it automatically detect the number of cpu on my dual 1800+ and it gets about 160-170k per client. Thanks again mathguy!

 

[InverseOfNeo]

Originally posted by: Baldy18
I let it automatically detect the number of cpu on my dual 1800+ and it gets about 160-170k per client. Thanks again mathguy!

Hmmm....thats pretty wierd that my dual P3-S 1.26GHz would have less than half of the production that dual 1800+ would get. Or am I completely wrong?

 

[MathGuy]

If I'm remembering correctly, the SoB client makes good use of the SSE2 extensions in the P4 and the AMD architecture seems to be generally advantageous for the client. Bottom line is that P3 < Athlon < P4, even for clock speeds that run about the same in most benchmarks. The client also appears to be rather heavily influenced by memory bandwidth (which doesn't help the P3 comparison either).

 

[InverseOfNeo]

Originally posted by: MathGuy
If I'm remembering correctly, the SoB client makes good use of the SSE2 extensions in the P4 and the AMD architecture seems to be generally advantageous for the client. Bottom line is that P3 < Athlon < P4, even for clock speeds that run about the same in most benchmarks. The client also appears to be rather heavily influenced by memory bandwidth (which doesn't help the P3 comparison either).

Thats really interesting. Although my dual system is on a DDR chipset (PC2100). Only dual Tualatin board that supports DDR.