Man, I am going insane. I recently purchased an MSI NVIDIA RTX 2060 Super. Most everything runs great but sometimes I get a random reboot. Nothing prompts it, it could happen in Excel or Photoshop or games. So I do a fresh install of Windows 10 and drivers et al.

The fresh install seemed to clear up a lot of the random reboots, excepts in games. Now I can't get more than 5 min into a game (any game) without a random reboot.

I don't know how to even diagnose this. Yes I have lots of RAM, but I have to uninstall the processor heat sink to remove any sticks of RAM so, I figured I would rule out the Graphics crad.

BTW, I have run stress tests on both the card and the CPU for hours and never had any issues or heat spikes.

Did you do a full uninstall of drivers, especially if you came from a different brand of card?

Do you have a different, working video card you can use to test with? Such as the old card you upgraded from?

(EDIT: I missed that you stress tested both and temps were fine.) Have you monitored your CPU/GPU temps? Sometimes, if you aren't running MSI Afterburner (or equivalent software, like EVGA Precision) the GPU fans won't run. The fact the crash time is delayed suggests it may be heat related.

Of course, you say that the stress tests work, which is really odd that they would work but a game wouldn't. Do you have an overclock setup?

If the old card works fine and the new card doesn't, I would call MSI support and set up the RMA right away.

If the same problems continue with the old card, then you know it is likely something else.

I had a really unstable system caused by dead Corsair watercooler pump that couldn't cool the processor, so the CPU would shutdown to protect itself. Replaced that cooler, and everything worked.

Does you MB have integrated video? If it works, given you fresh installed the OS, I wouldnt waste any more time and RMA.

I went to Microcenter and power supply was the same thought. So now, I've got a 1000 watt going in and I will update after that. My current 850w must not be enough for keeping up with all this beast is asking for.

AMD Threadripper 16-core, MSI RTX 2060 Super, 128GB RAM, 3 NVME drives and 3 HDDs.

if that 1000w doesnt work, I have a 1600w laying around for cheap. good luck

agfox06 said:

---

I went to Microcenter and power supply was the same thought. So now, I've got a 1000 watt going in and I will update after that. My current 850w must not be enough for keeping up with all this beast is asking for.

AMD Threadripper 16-core, MSI RTX 2060 Super, 128GB RAM, 3 NVME drives and 3 HDDs.

---

I have a Threadripper 1950X and a Vega 64 in my system. This runs without reboots on a 600W PSU. If I had to make a guess, I would bet something in your system is running too hot. If either your CPU or GPU hits 100C under load, then your system will shut off to prevent damage to the hardware. Threadrippers, with their large IHS, tend to be notoriously difficult to keep cool, so if you're using an air cooler that doesn't cover the whole heat spreader, or even a closed-loop liquid cooler that doesn't cover the whole IHS, then I'd imagine the Threadripper or GPU is getting too hot. Enermax (wouldn't recommend), Be Quiet!, and Noctua all make coolers for TR4 CPUs like Threadrippers that keep them cool relatively effectively. If the suspect is truly the GPU, you may just need to ensure that the GPU is able to push its heat out of the machine effectively. Most reference model GPUs force their heat out of the back of the card with a blower-style cooling solution, which works fairly well for forcing heat out of the system. Most custom cards, like MSi's, tend to have fans slapped on the side that do a good job getting heat out of the GPU, but don't account for forcing heat out of the system. Since this seems to have happened after you upgraded your GPU, you might have a situation where your GPU produces more heat, keeps it trapped inside the case, and then your Threadripper heats up faster and then hits the 100C cutoff. You could test this by putting your PC in an open-air area, then taking off your case's side panel, then playing your games at the settings that they are failing under currently. If your system doesn't fail in this configuration, you very likely have too much heat in the system and need to upgrade the CPU cooler or add more fans to get the hot air out of your case.