Motion Blur Crashing Scene

[juanjgon]

Yes, can be a hardware issue due to the overheating, or even due to a problem with the mainboard. Perhaps you should try to remove two of the four GPUs to get better cooling and see if the system is more stable. As I said before, if you have this problem with several scenes, that sounds to me like a hardware or system issue.

Thanks,
-Juanjo

[mikefrisk]

Just a thought:

Could it be a power issue?

With 4x gtx 1080ti i know realize my 1300 watt ps is probably too weak... I'm fairly certain i've got it to crash on only one gpu before but that could have been a fluke.

Just trying to nail down the source of this before I go and rma my mobo.

[juanjgon]

If the problem is that the power supply is not powerful enough for all the GPUs plus the whole system, the system should work fine with only one or two GPUs enabled. Of course, it is possible that the whole power supply is broken, but try first to render with only one or two GPUs to see what happens.

Thanks,
-Juanjo

[Fake Pilot]

@mikefrisk

I also have a brand new PC with 4 GTX 1080 Ti, experiencing a lot of errors of type 702, where it freezes my computer for a while.
Really interested if we can find this issue and solve it...

Code: Select all

CUDA error 702 on device 2: the launch timed out and was terminated
  -> failed to copy memory to device.
CUDA error 702 on device 2: the launch timed out and was terminated
  -> failed to launch kernel(init3)
device 2: direct light kernel failed
CUDA error 702 on device 1: the launch timed out and was terminated
  -> failed to copy memory to device.
device 1: failed to upload data texture 4 of context 0
CUDA error 702 on device 1: the launch timed out and was terminated
  -> failed to deallocate device memory
CUDA error 702 on device 1: the launch timed out and was terminated
  -> could not get memory info
CUDA error 702 on device 3: the launch timed out and was terminated
  -> failed to load symbol data to the device(imager data)
device 3: failed to upload imager params

[juanjgon]

If you have this problem with all your scenes, even with a simple one, the problem again sounds to me like a hardware or system issue. Try to enable the GPUs one by one to see if there is a problem in any of them. If the problem happens only when you have 3 or 4 GPUs enabled, the problem can be also the power supply.

Thanks,
-Juanjo

[mikefrisk]

I also have a brand new PC with 4 GTX 1080 Ti, experiencing a lot of errors of type 702, where it freezes my computer for a while.
Really interested if we can find this issue and solve it...

Which motherboard and powersupply do you have? I've ordered a 1600W power supply, when it arrives and I install it I will report back here. I'm thinking my computer is just under powered.

Like Juanjo said, the crashing happens only when 3+ GPU's are enabled. If it's not powersupply then maybe motherboard?

[Fake Pilot]

I have 1500W PSU and ASUS X99-E WS ECC-Ready Intel X99 Workstation ATX.

Did Octane Bench 3.06.2 test with all 4 separately, without a problem.

GPU 1: 179 – the one I use for 4K display
GPU 2: 193
GPU 3: 190
GPU 4: 193

All GPU together: 752. No problem.

I was running Test Version 3.07-Test3a for Cinema 4 and Octane Render 3.07 Test 3.
Maybe it's the plugin version?

[juanjgon]

Hmm, not sure, you are posting in the LightWave plugin forum

Ask in the C4D forum to see if there you can get more details

-Juanjo

[Fake Pilot]

Awh! Damn, I'm so sorry. Was searching for the same CUDA error.

But just for your info, installed 3.06.2 and got the same error as before. So not the C4D plugin version at least.
Could try another software.

[mikefrisk]

I have 1500W PSU and ASUS X99-E WS ECC-Ready Intel X99 Workstation ATX

I have the same mobo, the fact that you have a 1500W PSU is discouraging. With that information i'm now less confident its a PSU problem. Maybe the motherboard can't handle all four PCIe slots under load? If that is the case then damn... that would suck.