[beserkjay]

So I just got these

http://www.newegg.com/Product/Produc...82E16820231021

My question is, how can I test them without the possiblility of my CPU causing errors. I ran prime last night at 250x11 vcore at 1.55v and got errors after 4 hours (blend test). Could this be caused by my cpu?

How can I just test the memory without having the cpu influence wheither or not my ram is stable.

Thanks.

 

[Prelithe]

It's possible it's the cpu's fault, but before you check the cpu, download memtest and create a memtest boot floppy/cd and shove it in your drive and restart. Before you get into windows memtest will post up and test out your ram. People say a couple rounds of test #5, #6, #8 without errors should do the trick. Another way (or do both) to check RAM is using SuperPI (windows prog). Firstly run 1M, then if your alright then try 32M a couple of times. If you get problems, either run try running it on 2T (if you have it on 1T), relax the timings a bit (move CAS to 2.5) or even try running the mem on a divider (166 should be alright)

If everything is a/ok, then drop the HTTmulti down to 3x and/or drop the HTT freq down 5 mhz. That should solver the problem i hope. Also, check your temps, 1.55v seems a bit much, i have a 3200+ and can get it upto 2.65 at 1.475v with decent temps.

 

[dangerousHobo]

that RAM should be stable if its at 250mhz. You could turn your oc on our CPU down a bit and oc just the RAM and test it

 

[pbasil1]

download and run memtest for a few hours....

 

[beserkjay]

Alright I will try that tonight.

 

[Dopin_Nuts]

Yeah, do what prelithe said. And if you want, you can lower the multiplier down to 6-8 to ensure that the CPU isn't causing anything then run Prime95 again. It's probably your CPU, I have that same RAM and I can get it stable at 275. The only reason why your memory would be crapping out would be because you set the timings too tight, or you raised the vdimm over 2.6v. This memory is kinda weird, if you raise the vdimm above 2.6v it will actually cause instability. Also, for 250 you should be able to run your memory at 1T-3-4-8-4-12-14 pretty easily, but remember to keep the vdimm at 2.6v.

 

[Johnnie Walker]

You can also try running a divider with your ram running faster than your fsb, and crank it up and see if it is your ram that is not clocking properly.

 

[beserkjay]

Quote:

Originally Posted by Dopin_Nuts Yeah, do what prelithe said. And if you want, you can lower the multiplier down to 6-8 to ensure that the CPU isn't causing anything then run Prime95 again. It's probably your CPU, I have that same RAM and I can get it stable at 275. The only reason why your memory would be crapping out would be because you set the timings too tight, or you raised the vdimm over 2.6v. This memory is kinda weird, if you raise the vdimm above 2.6v it will actually cause instability. Also, for 250 you should be able to run your memory at 1T-3-4-8-4-12-14 pretty easily, but remember to keep the vdimm at 2.6v.

I got a general idea of the DRAM config from the guys over at DFI street. I can post the exact settings later.

When I first build my computer I could run 250x11 at 1.52v no problem, but ever since I upgraded my BIOS nothing I did could officially get it stable. I don't know what I did.

 

[Prelithe]

well if problems occured after bios flash then you should revert back to your default vers (after changing your settings back to before you flashed and still getting problems). You shouldn't really flash when you don't have any problems (in my book) unless it's a uber bios vers.

 

[beserkjay]

Quote:

Originally Posted by Prelithe well if problems occured after bios flash then you should revert back to your default vers (after changing your settings back to before you flashed and still getting problems). You shouldn't really flash when you don't have any problems (in my book) unless it's a uber bios vers.

Ya I know that now

But I already tried flashing back and my old settings were no longer stable.

 

[The Duke]

Set CPU to 7x to test the RAM.

 

[beserkjay]

Quote:

Originally Posted by The Duke Set CPU to 7x to test the RAM.

Should that work at stock volts?

 

[The Duke]

By underclocking the CPU that far it should not effect the OCing/testing of the RAM. I should run fine @ default at that speed

 

[beserkjay]

Quote:

Originally Posted by The Duke By underclocking the CPU that far it should not effect the OCing/testing of the RAM. I should run fine @ default at that speed

ok I will try that when get home as well.

But if they do turn out to be stable at 250x7 it doesn't necessarily mean that they will be stable at 250x11 right?

 

[Loki]

You dont need to download Memtest you already have it in your Bios...

 

[beserkjay]

Quote:

Originally Posted by Loki You dont need to download Memtest you already have it in your Bios...

Ya i know, I ran it for 3 passes when i first got the RAM but not overnight. I gamed for about 2 hours fine, but then got errors in prime after 4 hours

BTW is it possible for a CPU to require lower voltage after it has "burned in"? I may try lowering it to 1.5v and see if its stable at 250x11...

 

[The Duke]

If you read down about 1/2 way there is a quick test for memtest ...
When you think you have a stable speed run the full ver.
It should hit the same speed with differnet CPU multiplyers.
http://www.overclock.net/faqs/38268-...roduction.html

 

[beserkjay]

Update! This is what I was greeted to when I woke up so it looks like my RAMs fine

(blend test of course)

 

[beserkjay]

Ok so i tried this before I left for work this morning....(check pictures)

And I guess its not stable...any advice?

 

[beserkjay]

Ok so I tried bumping up the LDT and chipset voltage to 1.3V and 1.6V still didn't help. My LDT multiplyer is 3x.

Anything else I can try besides the cpu voltage?