Prime95 Blend freezes my computer.

[Ludo_Down]

Hard reboot required. I'm hoping this has something to do with my stability issues.
Is this indicating a memory failure? A CPU failure? How can I be more certain?
Memtest always runs fine if that helps. OCCT stress tests on the PSU, CPU and GPU do, too.

Thanks for reading, please help if you can. I am very inexperienced so even seemingly obvious info could help.

 

[unksol]

you are probably overheating the CPU. what are your load temperatures

 

[Anonymous_26]

I agree with unksol. Have you downloaded a temperature monitoring software like CPUID Hardware monititor? If so what are your temps reading?

 

[jimmysmitty]

A Prime95 blend test could be CPU or RAM as it uses both. It is probably due to an unstable CPU or memory setup.

Are you overclocking at all? As well what are your temps under Prime95 load?

 

[amuffin]

Temps?

 

[Ludo_Down]

I use HWMonitor to check all my temps. According to this, my CPU temps are never higher than high 70's low 80's C. That's during 3D graphics games.

 

[Ludo_Down]

And I'm not overclocking.

 

[Anonymous_26]

And it's just freezing, it's not giving you any BSOD?

 

[Ludo_Down]

Yeah just freezing during Prime95. It also freezes during other programs and loops last bit of sound. Never BSOD, never reboots on its own, just freezes in its tracks. Actually froze while I was writing this post.

 

[jimmysmitty]

High 70s to low 80c is pretty high. What CPU do you have?

Even on the stock cooler, Intel nor AMD should hit higher than 60cs when gaming since most games are GPU limited.

 

[Ludo_Down]

Well I'm goin from memory so let me test right now and give you exact numbers. This is my build:


CPU: AMD Athlon II X4 (645)
Mainboard: MSI 870-G45
RAM: 4x2 GB DDR3
GPU: AMD Radeon HD 6900
HDD: Seagate 2 TB
PSU: Corsair GS700W

 

[Ludo_Down]

Oh nm I was thining of the GPU temp. The CPU's max temp after 20 min or so of game was 37 C. This is usual I haven't seen it much higher.

 

[Ludo_Down]

Actually, though my temps are normal something did occur to me. It's a bit of an unrelated topic so I started a new thread:

http://www.tomshardware.com/forum/331005-28-damage-improper-thermal-paste-installation

 

[robertnewton]

Sounds to me like either an overheating / faulty CPU or that the power supply is too weak.

 

[4Ryan6]

RAM: 4x2 GB DDR3

Can you be more specific regarding your memory you're listing 4 2gb modules, are they a matched set?

What brand and model#s.

Drop back to 2 in the primary slots and see if the problem continues.

Sometimes leaving the memory on auto settings instead of manually setting the speed, timings, and voltage can cause problems when running 4 modules, especially if it's not a matched set, even when not overclocked.

 

[Ludo_Down]

Sorry I can't recall the exact model and the system in question is doing a malware scan so I can't get inside to check but I do know they are all matched and also I used to have just 2 sticks, problem existed then too. My plan tomorrow is to isolate each stick and test for crashes with just one (because they are matched I can't tell which are old or new) to see if there is a faulty stick among them (it would have to be one of the original two).

Any suggestions on a fast way to check four sticks individually?

 

[Ludo_Down]

FALSE: Just kidding. Shows what I know about my own system. My original sticks are two ADATA DDR3 (2GB each) and the two newer ones are Corsair XMS3 DDR3 (2GB each as well.) I'm currently isolating each stick and testing for crashes.

 

[Ludo_Down]

Can you be more specific regarding your memory you're listing 4 2gb modules, are they a matched set?

What brand and model#s.

Drop back to 2 in the primary slots and see if the problem continues.

Sometimes leaving the memory on auto settings instead of manually setting the speed, timings, and voltage can cause problems when running 4 modules, especially if it's not a matched set, even when not overclocked.

Good catch... since posting this I started testing my RAM sticks in isolation. They are NOT a matched set as I thought, and they seem to work well on their own but not together. I'll try 2 at a time when I get done with each one separately, but any specific instructions on exactly how to manually set the speed, timings, and voltage (and what to set them to) would be much appreciated.

 

[4Ryan6]

Good catch... since posting this I started testing my RAM sticks in isolation. They are NOT a matched set as I thought, and they seem to work well on their own but not together. I'll try 2 at a time when I get done with each one separately, but any specific instructions on exactly how to manually set the speed, timings, and voltage (and what to set them to) would be much appreciated.

Need Brand, Model#, of each set, the timings and voltage are usually on a sticker on the side of the modules?

 

[Ludo_Down]

Best answer selected by Ludo_Down.

 

[Ludo_Down]

Drop back to 2 in the primary slots and see if the problem continues.

Sometimes leaving the memory on auto settings instead of manually setting the speed, timings, and voltage can cause problems when running 4 modules, especially if it's not a matched set, even when not overclocked.

This seemed to be the issue, I ran just the Corsairs together in the 2 blue slots on my mobo and it didn't crash Prime95. Then I added the ADATA's to the black slots and no crash either. Strangest thing but the net change was really just that the ADATA's and the Corsair's switched sides (from blue to black and vice versa) but who knows maybe one was just slightly unseated or something. I am still hoping not to see a freeze on my system outside of Prime95 but if I do the first thing I'll try will be to remove the ADATA's. Thanks for all the help everyone!

 

[Anonymous_26]

High 70s to low 80c is pretty high. What CPU do you have?

Even on the stock cooler, Intel nor AMD should hit higher than 60cs when gaming since most games are GPU limited.

Yea but if he is running Prime 95 with a stock Intel or AMD cooler it can cause the temps to sky rocket pretty easily. I ran Prime 95 on a Phenom II with a stock heatsink and it only took about 20 seconds to hit 90+ degrees celcius.

 

[Ludo_Down]

High 70s to low 80c is pretty high. What CPU do you have?

Even on the stock cooler, Intel nor AMD should hit higher than 60cs when gaming since most games are GPU limited.

This was an error on my part anyway, that was GPU temps not the CPU. CPU was much lower even under heavy load.