The system reboots itself automatically when Prime95 is run

[EarthDog]

Room temps and system temperatures correlate 1:1. So if his testing was in a 19C environment as opposed to 25C, everything was, in general, 6C cooler. This difference could make the parts not flake out and reboot the PC. If a cutoff point is 110C and now it runs 105C.........

 

[Master]

Fair enough.

Make me a favor, stop using Prime95 and switch to AIDA64 run a stress CPU test and let's see if it crash if it doesn't crash run one again AIDA64 stress test but this time Stress FPU.

View attachment 92542

OK, I ran both tests for 2 hours each and hopefully nothing happened.
These are the latest temps before I stop the FPU test :

  DATE      TIME      MHz    CPU_0 CPU_1 CPU_2 CPU_3  LOAD%  GPU
09/28/17  21:52:50  3990.76    69    70    69    67   100.0   38
09/28/17  21:52:51  3990.76    69    71    74    70   100.0   38
09/28/17  21:52:52  3990.76    68    71    74    70   100.0   38
09/28/17  21:52:53  3990.76    69    72    74    69   100.0   38
09/28/17  21:52:54  3990.76    68    71    74    67   100.0   38
09/28/17  21:52:55  3990.76    69    71    74    69   100.0   38
09/28/17  21:52:56  3990.76    68    70    73    69   100.0   38
09/28/17  21:52:57  3990.76    71    72    74    70   100.0   38
09/28/17  21:52:58  3990.76    69    70    71    70   100.0   38
09/28/17  21:52:59  3990.76    67    72    72    69   100.0   38
09/28/17  21:53:00  3990.76    67    72    74    69   100.0   38

They are higher than when I was running the CPU stress (which was around 52~ 57 each)

 

[thebluebumblebee]

That's why I'm asking for help so that I can narrow it down to a specific part

Without some very expensive equipment, the only way to figure out what part is at fault is to use the process of elimination, which entails the swapping of parts.

Everyone's zeroing in on his PSU, and that could very well be true, but he doesn't appear to have a POS PSU:

There's only one 700 watt PSU from them, and it's the GP700A-HP, made by High Power. (the HP column) It *should* be an okay PSU. It is rated for full power at 40°C.

What I'd be concerned with:

• VRM's on the motherboard getting too hot. That motherboard may not have been designed with a tower cooler in mind. I would actually wonder if a top down HSF, like the Phanteks PH-TC14CS or the Noctua NH-C14S
  (reviewed here at TPU:https://www.techpowerup.com/reviews/Noctua/NH-C14S/) might be the better choice?
• Quality of electricity. IIRC, someone else was having problems with an ASUS motherboard and it was related to the "ASUS 5X Protection". If your power delivery is not very good, that may be causing problems. Worst case scenario, you'd have to get a line conditioning UPS.

OP, what fans do you have in your case and how they are configured?

 

[FireFox]

Without some very expensive equipment, the only way to figure out what part is at fault is to use the process of elimination, which entails the swapping of parts.

Everyone's zeroing in on his PSU, and that could very well be true, but he doesn't appear to have a POS PSU:

There's only one 700 watt PSU from them, and it's the GP700A-HP, made by High Power. (the HP column) It *should* be an okay PSU. It is rated for full power at 40°C.

What I'd be concerned with:

• VRM's on the motherboard getting too hot. That motherboard may not have been designed with a tower cooler in mind. I would actually wonder if a top down HSF, like the Phanteks PH-TC14CS or the Noctua NH-C14S
  (reviewed here at TPU:https://www.techpowerup.com/reviews/Noctua/NH-C14S/) might be the better choice?
• Quality of electricity. IIRC, someone else was having problems with an ASUS motherboard and it was related to the "ASUS 5X Protection". If your power delivery is not very good, that may be causing problems. Worst case scenario, you'd have to get a line conditioning UPS.

OP, what fans do you have in your case and how they are configured?

What i don't understand is that Prime95 makes the system crash and AIDA64 running FPU for 2 hours and the system didn't crash, dont tell me AIDA64 FPU it is not a heavy stress test.

 

[EarthDog]

These are the latest temps before I stop the FPU test :

I keep seeing.... the LATEST temps...what were your MAXIMUM temperatures??? P95 sorts through different length FFTs with different instructions etc and will change temps.

So, what was the MAXIMUM?

What i don't understand is that Prime95 makes the system crash and AIDA64 running FPU for 2 hours and the system didn't crash, don't tell me AIDA64 FPU it is not a heavy stress test.

It may not be AS heavy... and it isn't in fact over P95 in my experience.

 

[xorbe]

What i don't understand is that Prime95 makes the system crash and AIDA64 running FPU for 2 hours and the system didn't crash, dont tell me AIDA64 FPU it is not a heavy stress test.

CPUs are very complex, and sometimes it simply takes a certain pattern to tickle the troublesome spot.

 

[FireFox]

@Master

Did you try running an old version of P95?

 

[Master]

I keep seeing.... the LATEST temps...what were your MAXIMUM temperatures??? P95 sorts through different length FFTs with different instructions etc and will change temps.

So, what was the MAXIMUM?

It may not be AS heavy... and it isn't in fact over P95 in my experience.

Maximum was 76 C.

@Master

Did you try running an old version of P95?

No, I just grabbed the latest version.

 

[FireFox]

Maximum was 76 C.

No, I just grabbed the latest version.

Would you mind to try an old version?

 

[CAPSLOCKSTUCK]

So

after intensive testing the problem hasnt reoccurred?

 

[Master]

Would you mind to try an old version?

Which version do you suggest ? would any random version do it?

So

after intensive testing the problem hasnt reoccurred?

Nope, not yet.

Without some very expensive equipment, the only way to figure out what part is at fault is to use the process of elimination, which entails the swapping of parts.

Everyone's zeroing in on his PSU, and that could very well be true, but he doesn't appear to have a POS PSU:

There's only one 700 watt PSU from them, and it's the GP700A-HP, made by High Power. (the HP column) It *should* be an okay PSU. It is rated for full power at 40°C.

What I'd be concerned with:

• VRM's on the motherboard getting too hot. That motherboard may not have been designed with a tower cooler in mind. I would actually wonder if a top down HSF, like the Phanteks PH-TC14CS or the Noctua NH-C14S
  (reviewed here at TPU:https://www.techpowerup.com/reviews/Noctua/NH-C14S/) might be the better choice?
• Quality of electricity. IIRC, someone else was having problems with an ASUS motherboard and it was related to the "ASUS 5X Protection". If your power delivery is not very good, that may be causing problems. Worst case scenario, you'd have to get a line conditioning UPS.

OP, what fans do you have in your case and how they are configured?

I'm kinda getting the feeling the PSU is not the issue especially after reading this and this (since none of the symptoms mentioned there, apply to me). and as you theorized, it might be the VRM getting too hot.
I truly hope that it's not the Quality of electricity ! is there anything/any sign that I could check for and see if that's the case or not?

 

[FireFox]

Which version do you suggest ? would any random version do it?

27.7

 

[Master]

27.7

Cant find this version! but I could find 28.9 which seems to be from January.
By the way Aida64, Stress CPU/FU/GPU altogether ran for 1 hour without any issues as well.

 

[FireFox]

Cant find this version

http://www.mersenneforum.org/showthread.php?t=16779

 

[thebluebumblebee]

what fans do you have in your case and how they are configured?

????

I truly hope that it's not the Quality of electricity ! is there anything/any sign that I could check for and see if that's the case or not?

It is another hard thing to troubleshoot. Your reaction leads me to believe that you think you have good electrical service. This would be one of the last things we try to eliminate, unless you don't trust your electrical service.

 

[Master]

http://www.mersenneforum.org/showthread.php?t=16779

OK, 6 hours passed without any issues.

OK, I ran OCCT as well, despite some freezing (which would occur every several seconds) no reboot occurred. here are the images after 2 and a half hour of running OCCT.

????
It is another hard thing to troubleshoot. Your reaction leads me to believe that you think you have good electrical service. This would be one of the last things we try to eliminate, unless you don't trust your electrical service.

I guess now it's time to check this as well and see if thats the case