PC crashing - BSOD Watchdog violation 1x133 (possibly nVidia issue)

thanks a lot for your reply.

I will try the gpu hardware test...

If I'll continue to have issues seems there is no way to solve it properly?

If it's hardware, you can replace it once you know which part if faulty. If it's a software driver issue, you can ask NVIDIA to step in, find more users affected and ask them to contact NVIDIA, too, and wait until it's fixed.

Many of us (end users) had to wait for several months until the MSI fix became available by a third party and NVIDIA hasn't solved the problem (in the driver). I can only assume that NVIDIA doesn't want people to use 3 or 4 consumer-labeled GPUs for work when its high end products make lots more profit, but not every market can afford the high price tag of Quadros.

As a consequence of NVIDIA's practice, I've since built a second workstation using AMD's Vega Frontier Edition GPUs, to diversify hardware and software solutions (see OpenCL software, in place of CUDA) and I've stopped buying CUDA-only software. It's a loss-loss scenario, as NVIDIA loses sales and I loose in potential performance, but this is a well-known consequence of monopolistic behavior.

thanks a lot for your reply.

I will try the gpu hardware test...

If I'll continue to have issues seems there is no way to solve it properly?

I tried to apply with my 2 x rtx 2080ti stack I experienced more stability but still bsod occured randomly…

:(

Maybe I should change my Asus x299 W mobo?

Test your system with one GPU at a time. Can you test GPU1 and GPU2 on 2 PCI-E slots (e.g. PCI-E1 and PCI-E2) as follows:

1. GPU1 on PCI-E1, check for stability
2. GPU2 on PCI-E1, check for stability
3. GPU1 on PCI-E2, check for stability, and
4. GPU2 on PCI-E2, check for stability.

If one of the GPUs performs without any issues on both PCI-E slots and the other GPU has issues on both PCI-E slots, it's most likely a GPU hardware issue.

If both GPUs works fine in one PCI-E slot and cause BSOD in the other, I'd continue testing one GPU at a time in each of the remaining PCI-E slots, to ensure I've identified the defective PCI-E slot. If you identify a particular PCI-E slot that causes BSOD, it's most likely a motherboard hardware issue (e.g. bad/cold solder).

If all tests with one GPU at a time pass successfully, I wouldn't replace any hardware, I'd contact NVIDIA and ask them to look into the drivers, providing your hardware configuration.

In the past, I had an X99 mobo and an X299 mobo replaced twice (!) because of BSOD and because NVIDIA's tech support was blaming the motherboard and ASUS's tech support accepted it could be a motherboard issue (!!) and what I found was that it was NVIDIA's fault in the software drivers.

It's still unclear to me why NVIDIA recommends that we use a third-party utility to enable MSI, while NVIDIA could enable MSI upon installation of the software drivers.

I hope you find the above useful.

I tried to apply with my 2 x rtx 2080ti stack I experienced more stability but still bsod occured randomly…

:(

Maybe I should change my Asus x299 W mobo?

Did someone resolve that issues with asus x299 mobo and rtx nvidia driver?

thanks

The aforementioned fix (MSI enable) mitigates the DPC watchdog violation BSOD issue on my X299 Sage with 4x GTX 1080 Ti cards. I would expect this to be applicable to the RTX cards, too.