I posted this over at Intel CPU but I wanted to see if anyone with a P5B dlx has ran into this issue and hopefully overcame it.
I finally have the CPU watercooling in place and I've been doing some serious testing on the overclocking front. Here's what i've learned so far:
I can definitly attest to the issue described here:
http://www.thetechrepository.com/showthread.php?t=30
9x398 vs 9x404
-Inter-Core Bandwidth MB/s: 10527 vs. 10290
-Inter-Core Latency ns: 45 vs. 54
-Dhrystone ALU MIPS: 33011 vs. 33478
-Whetstone iSSE3 MFLOPs: 22704 vs. 22366
-Memory Latency ns: 68 vs. 84
-Speed Factor (lower is better): 80.7 vs. 98.2
-INT Buff'd iSSE2 MB/s: 7311 vs 6828
-Float Buff'd iSSE2 MB/s: 7333 vs 6783
The above stats are based off of 4-4-4-12-4 timing ( trfc 20, trrd 10, rank write to read: 10, read to precharge delay 10, write to precharge delay 10). Temps are in the 45C range.
I can't seem to get above 404 stable voltage wise. More vcore or fsb, nb,sb etc doesn't seem to help. Even to hit 404, I've got to up the anty quite a bit. Perhaps I should relax my timings to push beyond 404 but lLooking at that chart though, I'd have to hit like 455 to equal out at 398? I definitly don't think this setup can do that x9.
Am I going about this rightway? Reading the post above, if I drop the multi down to 8, it's going to give me probs with the nbcc.
9x398 for 3582 is a pretty decent return on this investment but it seems like i should be able to go a bit farther.
TIA,
Froz
I finally have the CPU watercooling in place and I've been doing some serious testing on the overclocking front. Here's what i've learned so far:
I can definitly attest to the issue described here:
http://www.thetechrepository.com/showthread.php?t=30
9x398 vs 9x404
-Inter-Core Bandwidth MB/s: 10527 vs. 10290
-Inter-Core Latency ns: 45 vs. 54
-Dhrystone ALU MIPS: 33011 vs. 33478
-Whetstone iSSE3 MFLOPs: 22704 vs. 22366
-Memory Latency ns: 68 vs. 84
-Speed Factor (lower is better): 80.7 vs. 98.2
-INT Buff'd iSSE2 MB/s: 7311 vs 6828
-Float Buff'd iSSE2 MB/s: 7333 vs 6783
The above stats are based off of 4-4-4-12-4 timing ( trfc 20, trrd 10, rank write to read: 10, read to precharge delay 10, write to precharge delay 10). Temps are in the 45C range.
I can't seem to get above 404 stable voltage wise. More vcore or fsb, nb,sb etc doesn't seem to help. Even to hit 404, I've got to up the anty quite a bit. Perhaps I should relax my timings to push beyond 404 but lLooking at that chart though, I'd have to hit like 455 to equal out at 398? I definitly don't think this setup can do that x9.
Am I going about this rightway? Reading the post above, if I drop the multi down to 8, it's going to give me probs with the nbcc.
9x398 for 3582 is a pretty decent return on this investment but it seems like i should be able to go a bit farther.
TIA,
Froz