PC Crashed during Intel BurnTest

[LFaWolf]

Hello,

I need some help with a strange problem that I have not experienced before. My previous motherboard Rampage IV Extreme went into a BIOS Update boot loop after I installed my new video cards, so I swapped out the board with a Rampage IV Black Edition that I have sitting around. Everything ran fine, but when I tried to test the stability @stock speed using Intel BurnTest, the machine crashed with BSOD (between running 30mins to 2 hours. One time it completed one loop before crashing again). I did not enable XMP or did any OC. I ran everything at stock speed to test stability first. I have CPU-Z, CPUID Hardware monitor, and Intel BurnTest running at the same time.

The specs:
Temperature- 57C load, <40C idle
-i7-3960x @stock clock
-CPU VCore 1.356 (Auto Setting, I tried to set it to 1.25 but for some reason it won't stay at that voltage).
-Noctua NH-D15
-GSkill RipjawsZ F3-17000CL11Q2-64GBZLD, capable of 2133MHz @1.5V but running @1333MHz SPD speed
-EVGA 1080 FTW in SLI (neither card ever exceeds 65C)
-EVGA G2 1300
-Corsair A540

The strange thing is I can game on it for a couple of hours, and 3D Mark Advance ran fine with Timespy, Firestrike Ultra/Extrme (CPU Physics), and Cinebench completed as well. I ran Prime95 for 2 hours but ran out of time because I had to leave.

Is Intel BurnTest still a good program to test stability? Should I just use Prime95? I am thinking doesn't matter which program, it should not crash @stock speed but I could be wrong? Could other background processes such as ASUS AI Probe or EVGA Precision X mess with the stability test?

Thanks in advance. I have been building PC and coding for almost 20 years but never have a machine to crash at stock speed. I have a theory but I want to hear from you what you think the problem lies.

 

[FR@NK]

Running 8 sticks of ram you might need to bump up the memory controller voltages.

 

[LFaWolf]

Is that what it is called VCCSA Voltage? Also forgot to mention that I ran memTest86 (latest version) @1333MHz speed and it completed without any errors.

 

[TheoneandonlyMrK]

Hello,

I need some help with a strange problem that I have not experienced before. My previous motherboard Rampage IV Extreme went into a BIOS Update boot loop after I installed my new video cards, so I swapped out the board with a Rampage IV Black Edition that I have sitting around. Everything ran fine, but when I tried to test the stability @stock speed using Intel BurnTest, the machine crashed with BSOD (between running 30mins to 2 hours. One time it completed one loop before crashing again). I did not enable XMP or did any OC. I ran everything at stock speed to test stability first. I have CPU-Z, CPUID Hardware monitor, and Intel BurnTest running at the same time.

The specs:
Temperature- 57C load, <40C idle
-i7-3960x @stock clock
-CPU VCore 1.356 (Auto Setting, I tried to set it to 1.25 but for some reason it won't stay at that voltage).
-Noctua NH-D15
-GSkill RipjawsZ F3-17000CL11Q2-64GBZLD, capable of 2133MHz @1.5V but running @1333MHz SPD speed
-EVGA 1080 FTW in SLI (neither card ever exceeds 65C)
-EVGA G2 1300
-Corsair A540

The strange thing is I can game on it for a couple of hours, and 3D Mark Advance ran fine with Timespy, Firestrike Ultra/Extrme (CPU Physics), and Cinebench completed as well. I ran Prime95 for 2 hours but ran out of time because I had to leave.

Is Intel BurnTest still a good program to test stability? Should I just use Prime95? I am thinking doesn't matter which program, it should not crash @stock speed but I could be wrong? Could other background processes such as ASUS AI Probe or EVGA Precision X mess with the stability test?

Thanks in advance. I have been building PC and coding for almost 20 years but never have a machine to crash at stock speed. I have a theory but I want to hear from you what you think the problem lies.

That much testing and stability isn't necessary, in all likelihood the imc or another bus is going flackey when it gets heat soaked or and through thermal hotspots, if you have replaced the Tim and cleaned heatsinks I'd still be ok if it was stable in use ,if you have not I'd start there.
Not many companies recommend the use of such software as they are counted as power viruses and not representative of any typical load too ,I would not be to bothered but imho a better test would be to join the world community grid crunching team here and run that a few days imho it's a better stability test and with hwinfo64 running and logging in the background plus the event viewer faults can possibly show easier.

 

[Mussels]

VRM'sd overheating on the board? people forget that IBT cooks them too. point a fan directly at them and leave it for a few hours, see if it helps.

 

[Jetster]

Don't use Intel Burn test. Nothing works your system that hard

 

[Mussels]

Don't use Intel Burn test. Nothing works your system that hard

if a system cant handle IBT, its not stable, even at stock.

 

[LFaWolf]

VRM'sd overheating on the board? people forget that IBT cooks them too. point a fan directly at them and leave it for a few hours, see if it helps.

Hmm, but I have the 2 CPU fans already cooling that area, unless you are talking about the chipset. My previous RIVE does have a fan to cool the chipset heatsink but for this board it is passive. I am actually thinking maybe the CPU is dying? Is that possible? I have never had a CPU died on me before, not to mention such a lightly used CPU (computer is only used for gaming). I did OC it previously to 43 then down to 4.1

 

[Arctucas]

Don't use Intel Burn test. Nothing works your system that hard

Well...

P95 and LinX do.

I can pass 3 hours of IBT at Maximum using my 24/7 settings, but almost immediately have at least one worker fail in P95 running small FFT, and LinX is even worse, if using the largest problem size.

 

[LFaWolf]

So how many hours of running Prime95 can be considered as stable? 6 hours? Should I forgo IBT and go with P95?

 

[Mussels]

So how many hours of running Prime95 can be considered as stable? 6 hours? Should I forgo IBT and go with P95?

you should try all available stress tests for a few hours each. some setups crash faster on different programs, so you find the one you're the least stable at and focus on it

 

[Arctucas]

So how many hours of running Prime95 can be considered as stable? 6 hours? Should I forgo IBT and go with P95?

I am sure you will receive different answers, but I believe a 24-hour run would indicate a very stable rig.

However, personally, I never use P95 for more than an hour.

My personal stability testing regimen includes:

3 hours IBT at Max.

8 hours of RealBench using all RAM.

At least 12 hours HCI Memtest.

8 hours X264 HD benchmark.

You should keep the CPU temperature under 80°C.