Cannot get past 250mhz with OCZ 2gb PC4000 Gold GX XTC kit

[polobunny]

I'll probably be sending my memory on RMA. Can't seem to get past 250mhz at "stock" timings.

I know 250mhz is what they say it should run @, but i've seen far better than this (265mhz 1T and even more in 2T from others) from Samsung UCCC chips. I also know the date code has alot to do with UCCC chips.

At first I was wondering if it wasn't my cpu being bulky so I've put a good 2.65v in my nF4 and 2.8v in my OCZ (everyone recommends 2.8v for them btw), applied a 3x HTT (duh!) and lowered the cpu/mem ratio from 1:1 to 3:2 (200/133, for the noobs :P). With 1.4v vcore I got up to 11x257= 2.825ghz prime stable. It rebooted at 2.85 (11x259), but lol it's alright this way I think it's not that bulky. I tried a few different multiplier 9, 9.5, 10, 10.5 and it always give the same results or about.
Whenever my ram goes over 250mhz, it fails. Really, I can be at DDR504 and it won't be prime stable and a DDR506 is will reboot instantly. Actually If I remember properly once it wouldn't make "9x298somethingish" @ 6:5 cpu/mem ratio 1.4v vcore thus making it NOT even able to reach 250mhz prime stable.

So yeah no matter what I do, the ram seems to be "limiting me"... :?

Does anyone have any other idea of tests I could try before sending this memory on RMA? Since my CPU seems to be fine with the tests I made. Maybe I forgot something?
Anyway I have my 2 good sticks of Samsung OEM 512mb TCCC's if someone think I should test my CPU with different memory. I don't have any other mobo at hand, and mostly no other CPU (I could always get a hold of this ultra-D but meh). I don't want to check my stepping I would need to remove my Big Typhoon.


Thanks for your input.

[gvblake22]

What have you been using to test for stability? Have you tried Memtest86+?

[polobunny]

No, only been priming it. I guess I should test it with memtest since prime can be pretty weird sometimes.

[Capper]

I'm not a big fan of prime. I know you'll hear a lot both ways, but Prime was developed for Intel processors....and I don't feel its a good way to test for memory stability on an AMD system

[polobunny]

Instant reboots don't lie though. :P

[Capper]

Doesn't hurt to test with different components to isolate the issue

[RangerXLT8]

Set everything else to default, rum memory w\ a divider to achieve desired speed. Then run memtest86 off a bootable CD. It's easy to do if you only have one variable(memory), everything else being stock.

[polobunny]

Can't do that on my board.
The bios is broken. Setting divider cpu/mem to 4:5 in bios makes the memory run at 166mhz or something XD
And no, I can't be wrong about that, as they clearly write "250mhz" setting and not 4:5.

[RangerXLT8]

At any rate you need to isolate the componet throwing errors. If you must use P95, FIRST run small FFTs(they only use the CPU cache, very little memory) for a few hr to make sure the CPU is stable. After that completes, then run memtest86 on your memory. There are better test out there for memory then P95, Prime95 is primarily meant for CPU. If you'r getting errors in P95, chances are it's the CPU and not memory.

[polobunny]

Well I ran prime with small FFTs for a while no error with my memory under 250mhz w/o a problem. When I get over 250mhz it takes less than a minute for an error to pop out. Anyway i'll check with memtest... sometime...soon. :P I don't want to do it now meh.