Fsb and multi

cje · 20 Feb 2009 at 17:42

Sorry to open a new thread, but I have burning question....

If I could get my striker II Extreme stable at 1600 fsb but with the cpu multi on 7x. Then, if I raised multi to 8 or 9, the only factor to change is vcore?

Or is it more complex than that. Because I cant for the life of me get this striker II extreme stable at anything above 1333 fsb. I've got a copy of what AndyOCUK had his rig to run a Q6600 at 3.6ghz, and had the same ram and density as I do.

I've got it 1600 with a 7x multi now, with the processor running 2800 mhz. I know it can run just fine at 3ghz with 1.27vcore. So the processor should be stable at 2800 with 1.3v.

It passes ALL linpack testing, but will bsod when playing some tunes in itunes after a while. My ram is at stock. I have set the command rate to 2T, Im now testing with 1T to see if it makes a difference.

Any help would be appreciated. thanks. :cool:

It's stable linpack and prime95 smallffts. Stable prime blend for 1 hours. It always bsods on idle. I've tried 1T and 2T for ram. Any suggestions?

w3bbo · 21 Feb 2009 at 00:13

First off run your CPU with the 9x multi. There really is no need to run it with a lower multi on an nvidia chipset as the FSB and memory is 'unlinked', or can be if you set it up to be. This will put less stress on the NB which could be the cause of your BSOD.

Set you CPU vcore to 1.5v
Set your FSB to 1600 (400mhz effective)
Set your CPU multi to 9
Set your DDR3 to its rated speed and voltage.

It might not be stable at this speed (3.6ghz) but it should boot and give an indication of if your chip is capable or not.

cje · 21 Feb 2009 at 00:34

Before you posted this, I have passed 2 IBT runs of 200 each with an 8x multi. I found my instability was due to ram. I set voltage to 1.5v (as it is stickered) and command rate to 2t as I've seen via the search function, and found andy's s2e thread.

Yeah, just set it to 1.5vcore in BIOS. But temps do get very toasty with Intel Burn Test. I let it pass one test, but temps shot up to 75, so instantly stopped it. It's stable at 3.2ghz with 1.35vcore in bios, but I guess 3.6 is too much of a jump heat wise.

cje · 21 Feb 2009 at 01:24

I lied. I left it at the 8x multi where it passed 200 IBT. I left it for 10 minutes and it had restarted itself, and frozen on boot! Nothing above 1333 fsb seems to work.

My ram is rated 9 9 9 24 at 1.5v. I've tried 1T and 2T no cigar. Seems to be a memory error. No errors when 1333fsb though. Mobo problem?

w3bbo · 21 Feb 2009 at 01:56

Are you locking your memory (Unlinked) to 1333FSB if not do this. If your memory is set to be linked/synced then it will be running at 1600mhz.

Have you tried running your memory at 1.6v?

cje · 21 Feb 2009 at 09:11

Thanks for the replies mate.

Yes I have put my RAM to unlinked. Set Timing manually to 9-9-9-24. I'll try 1.6v today, see if it'll be happy with that.

cje · 21 Feb 2009 at 12:46

Nope, didn't like that either. The RAM should be stable, because with the same timings and command rate and voltage is totally stable with 1333 fsb. Maybe my NB is the problem? I'll have a play with NB GTLs see if they help...

cje · 21 Feb 2009 at 21:46

Disregard what I said.
System BSODS and restarts after a while of linpack testing with the following settings:
Fsb:1600
Multi:8x
RAM:1333 (9-9-9-24 2T)

Vcore:1.4
RAM : 1.6
NB : 1.44

I tried it with 1T as well, no go.

Then I decided to put all voltages on auto, except vcore.
Vcore: 1.4
Vdimm : AUTO (Actual :1.66v)
NB : AUTO (Actual : 1.32v)

On 2T command rate it stayed stable on linpack for 4 hours, before deciding to restart. This is a much longer period before resetting.

I think it may be a RAM issue, but when I set the VDIMM manually to 1.6, the board overvolts to 1.66 anyway.

I am at a loss with this board by now. I'll try setting RAM voltage to 1.62v.

Any ideas?

EDIT - It seems to be unstable on idle. Passes x amount of linpack and x amount of hours of prime, but when the testing finishes itll bsod and restart. Or, it BSOD as I restarted it just now.

cje · 22 Feb 2009 at 18:38

Forget whatever I posted earlier. It is now 200 passes Intel Burn Test stable at 3.2 ghhz. I had to set NB GTL to +10mv in order to find stability. It would pass the tests for an hour, before bsod and a restart. The increase in NB GTL completely eliminated that.

Now for 3.6 ghz. I am not going to use Intel Burn Test for prolonged 3.6 ghz testing. My experience is that if it passes around 10 tests, then the cpu is more or less stable, but your mobo or ram might not be. As I found out. When my cpu is undervolted it'l restart in the first or second test in linpack, but if ram/nb is faulty, it might do it in the 80th or 100th test.