Low multiplier -> higher stable speeds??

[exscape]

OK, I posted this on Xtremesystems as well, thought I might post it here as well... Anyway, in case you're too lazy to even click the link:

 

I can't figure out what the hell is going on here! My CPU seems to be more stable at lower multipliers / higher HTT's.

 

Vcore @ 1.34v for all tests (don't ask )

9*289 = ~2600: not stable (1-2 minutes prime95)

8*323 = ~2600: not stable (same as above)

7*370 = ~2590: stable for 15 hours prime95

 

I'm hoping that some BIOS settings might be the the cause of this, so that I can do something about it!

 

Hardware in sig.

 

Two screenshots... Note the Vcore in both shots!

http://web.telia.com/~u18411604/venice/7x370.png

http://web.telia.com/~u18411604/venice/9x289.png

 

Any suggestions? I've tried fooling around with memory timings, not that I know why it would help, but it didn't, at least not before I gave up.

 

Edit: I might add that while 7x370 seems completely stable at 1.33-1.34v 9x289 fails prime95 within 1 minute every time.

[mad mikee]

Higher HTT = lower BW because of lower multi x whatever FSB you've divided into, so less stress on the CPU mem controller (unless that is 1:1, in which case you should send ME that memory to see whats wrong, but I'm sure testing will take YEARS :nod: :shake: )

[exscape]

I'm running it all on the DDR100 divider just to be sure... I don't see how it would matter, though... Shouldn't the mem controller have to do MORE work at 370/2 than at 280/2?

[DeepBlue]

did you try any other divider?

 

if not, maybe it's just that your ram hates such low of a speed (?)

[exscape]

Yup I've tried other dividers as well, doesn't help at all.

I gotta say that this is really annoying... Either I max my CPU at 9x289 with lots of vcore or I run at 7x370 with very little vcore but instead I can't get the CPU higher anyway because the board doesn't run with 375+ HTT

[Tom Brohanks]

Unless you're running 1:1, you gain no benefit from having a low multiplier and high FSB.

 

This has been proved. 2600mhz at 300fsb is the same as 2600 at 200fsb if you are NOT running 1:1

[exscape]

I know... But my CPU can run 2.6GHz @ 1.33v with 7x multi and 2.6GHz @ 1.46v (!) with 8x/9x multi.

I want to be able to get that nice and high speed (or more) using 9x

[Surrept]

only a 64 user would complain about their fsb being to high, almost the same thing as people complaining about their board being to cold

[exscape]

I'm still not *complaining* that it's too high. I just want it to be possible to achieve the same speed, at the same voltage, with a higher multiplier, since 371-373 seems to be the max for my board...

Seeing how 2600 MHz is stable at 1.34v I suppose 2700 isn't unreachable?

[Surrept]

poke it with a stick and i'd bet 2800 will be a breeze