Assimilator1
Elite Member
- Nov 4, 1999
- 24,125
- 508
- 126
lol 'congealed' ,I assume you meant frozenOriginally posted by: Ev3rs0n
I have stable 100% as well as this 1550-775, but when I put 1600-800 to him is congealed the Windows, I also apply but voltages to him to the Vcore and to the SB, but it is congealed.
I just recently built my P6N Plat rig & have only just started to overclock it significantly.
Firstly I hit a FSB wall at ~1500MHz ,this turned out to be the Northbridge getting too hot (I didn't have a fan fitted to it).Using my laser thermometer I measured the HS at 58C IIRC (which was just bareable to touch).
I didn't want to use that crappy fan MSI supplied (are they nuts revving it at 8500RPM!!??:Q:roll: ) so I suspended a 70mm fan a little way over it & it runs much cooler ,I just retested it earlier & it now runs at 42C @ 1625 MHz ,1.35v.
I haven't properly stability tested yet but I found that at 1550QDR (375MHz) I had to increase NBv to 1.275 (from def 1.25),
1575MHz upped to 1.3v ,
1600 upped to 1.325 ,
1625MHz upped to 1.35v.
Whether that'll end up being stable in proper testing is another matter ,I'll find out soon enough ,but it at least it shows the trend in voltage needs for FSB.
Oh & btw even at 3.25 GHz I haven't yet had to increase vcore on my E6420 (subject to proper testing)
I would like to increase NBv to 1.375 & maybe 1.4v to get higher FSB ,but I'm somewhat uneasy about this as even at 1.35v that's 8% over default.
What's considered a safe limit & how would we know? ,bearing in mind I could keep this mbrd for as long as 3yrs (& I want it to be saleable afterwards) & that I run DPAD on it 24/7 on both cores. (DPAD's a Distributed Computing project).
I'll update when I've chosen the optimal speed.