Question New Build - Restarts on 1-2 Minute of FurMark Stress Test

[_Makky_]

Hey everyone
I just made this build for a friend of mine and for some reason this system would restart of ~5-10 minutes of gaming or 1-2 minutes of stress testing.

My guess: Temperature / Power / Drivers / Hardware
What I have done so far:

1. Temp: FurMark CPU stress testing goes well, highest CPU temp after 10 mins of testing is ~73°C ~74°C
2. Temp: FurMark GPU stress testing initiates a restart within 1-2 minutes. The restart happens almost immediately when the GPU reaches ~72°C. Last info I see from FullMark GPU stress testing is ~72°C GPU temp and 70% fan RPM capacity.
3. Power: 550 Watt 80+ power supply shouldn't be an issue I guess coz everything is running at stock. However, I am not ruling out the possibility of insufficient power.
4. Drivers: Drivers were downloaded via IObit Driver installer initially. I have also tried downloading & installing drivers manually via AMD's website.
5. Hardware: I suspect GPU was plugged in before. Relevant Reddit post. GPU BIOS version: 015.050.002.001.000000. A quick check on techpowerup.com indicated the BIOS is stock.
6. BIOS: BIOS was already updated to latest 1607 version out of the box.
   
   
   The system sends no data to monitor and turn blank screen. The system freezes (numlock not responding) for a while before I have to forcefully restart the system.
   I couldn't find a way to look at BIOS logs to look for thermal trips. Is there such a feature in B450M-A ?
   
   
   I hope I have a lead and it's just a bad GPU and not bad config. I am freaking out as I have previously build 5-6 low spec systems and never encountered any issue. This my first gaming build and it's not going as planned. I really need some help 🙁
   

Specs:

Part​	Model No.​
CPU	AMD - Ryzen 5 3600
Motherboard​	Asus - PRIME B450M-A Micro ATX​
Memory​	Corsair - Vengeance LPX 8GB DDR4 3000MHz C16 (CMK8GX4M1D3000C16)​
Storage​	Samsung - 970 Evo Plus 250 GB M.2-2280​
Storage​	Western Digital - Caviar Blue 1 TB 3.5" 7200 RPM​
Video Card​	Sapphire - Radeon RX 580 8 GB Pulse​
Power Supply​	Cooler Master - MWE 550 W 80+​
OS​	Window 10 64 bit​

 

[ConanLock]

Did you wipe Windows before using the official drivers? If not, I'd recommend just wiping it and reinstalling. Driver downloaders can mess things up beyond the point just reinstalling the drivers fixes them. Also, do you use any system optimisers, speed boosters etc? If so, do not install them on the new Windows install.

 

[_Makky_]

Did you wipe Windows before using the official drivers? If not, I'd recommend just wiping it and reinstalling. Driver downloaders can mess things up beyond the point just reinstalling the drivers fixes them. Also, do you use any system optimizers, speed boosters, etc? If so, do not install them on the new Windows install.

Thank you for responding, just getting a response is assuring.

No. I did not wipe Windows before using the official drivers. I am gonna try that.
I do not use any system optimizers etc. Also, the system is fresh; no old components/parts installed.

 

[jeremyj_83]

Thank you for responding, just getting a response is assuring.

No. I did not wipe Windows before using the official drivers. I am gonna try that.
I do not use any system optimizers etc. Also, the system is fresh; no old components/parts installed.

I would want my CPU temps to be a bit lower than that, but it is still within spec.
My best guess is either drivers as ConanLock said or your PSU. The fastest to test is drivers and doesn't cost anything. The PSU is also a good chance to be the issue since the MWE is of low quality. While put under stress, it could be giving poor power delivery to the system. A good choice for a different PSU if it is needed is the Seasonic Focus Gold or Seasonic Focus Plus Gold. Both are of very good quality and have far higher efficiency than your current PSU.

 

[_Makky_]

Here's what I have tried as suggested by ConanLock:

1. Fresh Windows 10 64 bit installed.
2. Installed drivers for GPU, motherboard from the official source.
3. Installed FurMark GPU stress test.
4. GPU temp shoots from 50 C to 73 C in 1 Minute 20 seconds during the test.
5. Monitor goes off.
6. System freezes (numlock not responding), GPU, CPU and cabinet fans still spinning.
7. System restarts after some 20 minutes.

Still confused as hell. I haven't closed the cabinet cover yet so case air warming isn't the issue.

I had left the system running last night for 8 hours (while the cabinet was closed, playing some video on a loop overnight) and the system did not fail. CPU temp was 43 C and GPU temp was ~38 C in the morning.

Interestingly I have also tried running CPU and GPU stress test simultaneously. The system won't fail until GPU reaches 73 C. << This test makes me imply that this is not a PSU shortage issue as the watt drawdown will reach close to peak the moment both CPU and GPU are under 100% load. I assume 10-20% change in GPU fan RPM capacity isn't gonna draw significant load from PSU.

 

[jeremyj_83]

Here's what I have tried as suggested by ConanLock:

1. Fresh Windows 10 64 bit installed.
2. Installed drivers for GPU, motherboard from the official source.
3. Installed FurMark GPU stress test.
4. GPU temp shoots from 50 C to 73 C in 1 Minute 20 seconds during the test.
5. Monitor goes off.
6. System freezes (numlock not responding), GPU, CPU and cabinet fans still spinning.
7. System restarts after some 20 minutes.

Still confused as hell. I haven't closed the cabinet cover yet so case air warming isn't the issue.

I had left the system running last night for 8 hours (while the cabinet was closed, playing some video on a loop overnight) and the system did not fail. CPU temp was 43 C and GPU temp was ~38 C in the morning.

Interestingly I have also tried running CPU and GPU stress test simultaneously. The system won't fail until GPU reaches 73 C. << This test makes me imply that this is not a PSU shortage issue as the watt drawdown will reach close to peak the moment both CPU and GPU are under 100% load. I assume 10-20% change in GPU fan RPM capacity isn't gonna draw significant load from PSU.

Best guess based on that information is it is heat related.

 

[_Makky_]

Best guess based on that information is it is heat related.

So no faulty GPU?

How do I test if this is a heating issue? Also, should heating be an issue when everything is running at stock?

 

[_Makky_]

https://outervision.com/b/421YKY

This PSU calculator indicates:

Load Wattage:​	385 W	+3.3V​ +5V​ +12V​ 9.5 A​ 7.9 A​ 26.1 A​ 71 W​ 314 W​
Recommended UPS rating:​	800 VA
Recommended PSU Wattage:​	435 W

I guess CM 550 MWE with 80+ certification would work at least 80% efficiency and highest load would supply 440 W power.