• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Prime95 Blend instant bluescreen, other tests pass after overclocking

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

CyberClock

Registered
Joined
Apr 4, 2015
For some reason, only the blend Prime95 displays a bluescreen almost immediately, while the rest run fine. At stock settings the BSOD does not appear, but it does after overclocking it. My memory is fine, I tested two different memories and it still appears. I am using a brand new, error-free RAM. Here's the RAM I am using: http://www.newegg.com/Product/Product.aspx?Item=N82E16820226191 (no Memtest issues)

I have attached my settings below:


What could be causing the blend test to fail? I have adjusted the NB frequency up/down even under the stock settings and it still appears.
 

Attachments

  • 4KViwH1.png
    4KViwH1.png
    33.9 KB · Views: 1,788
Last edited:
If this is the only test it is failing on and doesn't happen during normal use, I wouldn't worry about it.

That said, we need more information...

Can you post CPUz screenshots. The first tab, memory, and SPD please?
 
Sure. As a side note, I upped the NB frequency just now. Has nothing to do with the current issue.

9iAYWE0.png
 
Look at your CPUz screenshots, notice how on the memory tab you are at 800Mhz(1600Mhz effective), and the timings are 9-9-9-24? Then look at the JEDEC #6 there... for 800MHz it shows 11-11-11-28. That would be the reason right there. Im not sure the link you provided in the first post are the same thing as those are, as you have it set, 9-9-9-24. What does the sticker on the actual ram say itself?
You look to own these: http://www.newegg.com/Product/Produ...572&cm_re=Mushkin_2133-_-20-226-572-_-Product

See how the P/N matches the SPD tab and the P/N at newegg?

I bet those errors would go away if the timings were set properly tothe 11-11-11-28 (but holy cow are those timings HIGH!).
 
Last edited:
If you do Small FFTs in Prime95 does it BSOD?

What's the BSOD error code?
 
Look at your CPUz screenshots, notice how on the memory tab you are at 800Mhz(1600Mhz effective), and the timings are 9-9-9-24? Then look at the JEDEC #6 there... for 800MHz it shows 11-11-11-28. That would be the reason right there. Im not sure the link you provided in the first post are the same thing as those are, as you have it set, 9-9-9-24. What does the sticker on the actual ram say itself?
You look to own these: http://www.newegg.com/Product/Produ...572&cm_re=Mushkin_2133-_-20-226-572-_-Product

See how the P/N matches the SPD tab and the P/N at newegg?

I bet those errors would go away if the timings were set properly tothe 11-11-11-28 (but holy cow are those timings HIGH!).

I set the timings to 11-11-11-28 and it did delay the BSOD for about 15-20 seconds, as opposed to near-instant.
If you do Small FFTs in Prime95 does it BSOD?

What's the BSOD error code?

No BSOD for other tests.
 
If you set the timings to 11-11-11-28 and you're still getting bluescreens it is likely you're Cpu Oc isn't stable. How did you get to the 3600 Mhz Oc. Did you just plug in numbers and try running prime? If you did you may want to start from the beginning and test it slowly. Drop the other multipliers Ie Ht Link, Nb Freq and Dram add 5 mhz to the Fsb test with prime blend and add voltage when you fail.
 
BSOD code?

Does it do this at stock speeds?

No code.

If you set the timings to 11-11-11-28 and you're still getting bluescreens it is likely you're Cpu Oc isn't stable. How did you get to the 3600 Mhz Oc. Did you just plug in numbers and try running prime? If you did you may want to start from the beginning and test it slowly. Drop the other multipliers Ie Ht Link, Nb Freq and Dram add 5 mhz to the Fsb test with prime blend and add voltage when you fail.

Yes, and I will try the advice of dropping other multipliers.
 
Just to be clear, load up default settings, then drop the other multipliers and then start testing using the FSB. The problem with just putting in settings that someone else has used is, every cpu is different. Very few will take the same voltage for X Oc as another will. You may also want to read this guide it can also be helpful.


ED maybe we can move this to the AMD cpu section?
 
Try changing command rate to 2T and run 11-11-11. Probably won't hurt to manually input ram voltage too.
 
Running IBT could pass the bench like a dream. Run P95 Blend mode and nearly immediate fail. This has been something that has happened to me more times than can count. When I finally would get P95 Blend to run it was when I reduced the max cpu clock (or more cpu voltage) or got the ram set up just right. Such strangeness used to boggle my mind and why I use both tests still today when passing my final 24/7 overclock as assumed stable.

RGone...
 
Well I accidentally destroyed my motherboard (presumably). I plugged in a really high number while overclocking my GPU (460 GTX) as opposed to incrementally increasing it, tested it with Furmark and as the temps were rising the computer instantly turned off with the lights blinking for 15 minutes before ceasing permanently. The lights never blink now.

When I press the power button, nothing happens. No fans, nothing. However, when I plug in a wireless adapter USB into my ceased computer, the green light is on for the USB. Does that mean it is just my motherboard and my PSU is working fine? My intentions is to buy a new motherboard and an upgraded CPU.

This is my current 700 Watt PSU: http://www.newegg.com/Product/Product.aspx?Item=N82E16817341018
 
Last edited:
Usually GPU overclocking doesn't kill motherboards. If you are not getting any power anywhere, that tells me its the PSU. Those OCZ ModExtreme's were "MEH" at best IIRC. Do you have another power supply to test with? That would be the first thing I do...
 
Video card could have drawn too much current thru PCI-e slot killing it and then maybe the motherboard. I agree the P/S needs to be jump started and voltages checked to see if they are within spec at least, even if not loaded.

RGone...
 
I will try jumping the PSU and report back.
Video card could have drawn too much current thru PCI-e slot killing it and then maybe the motherboard. I agree the P/S needs to be jump started and voltages checked to see if they are within spec at least, even if not loaded.

RGone...
Do you mean killing the PCI-e slot and motherboard, or video card and motherboard?
 
Regardless, you can try switching the GPU to another slot to see if it killed the slot. You can test the GPU in another PC to make sure the GPU didn't die...
 
Back