Analytic Lights & Post Processing crashing Octane Server
Posted: Sun Nov 09, 2025 11:44 pm
Big issue with Analytic lights causing some kind of memory error under certain situations. This is happening in Windows version of OctaneBlender and reproduced on two different computers.
I am using an Analytic light set to disc. Octane Server crashes while using Analytic lights with post processing in viewport preview render. The first initiation of viewport render sometimes works fine - one can see the post process effects on the Analytic lights, adjust settings, and freely move around during viewport rendering. However, if one shuts down rendered view by going back into shaded view and after that again attempts a viewport render - server crashes with following error: "ERROR! check console for detailed error message."
Various Errors from Console logs on numerous separate crashes while testing show this:
The error doesn’t occur on Mac. And I have only tried this with the current stable standard plugin not the add on beta. OctaneBlender version is 2025.3 - 30.7 [STABLE].
Can provide more detailed logs and specs of computers if necessary. Basically, I'm on Windows 10 using a 5090 and a 3090 on main machine and a 3080, 3080 Ti, and a 1080 Ti on the other.
NVIDIA Studio driver 581.29. Just updated to Studio driver 581.57 and issue persists (perhaps even more unstable).
The Mac, which tested fine, is a Mac Mini M1 running Sequoia 15.7.1.
Will post this in Discord channel as well.
I am using an Analytic light set to disc. Octane Server crashes while using Analytic lights with post processing in viewport preview render. The first initiation of viewport render sometimes works fine - one can see the post process effects on the Analytic lights, adjust settings, and freely move around during viewport rendering. However, if one shuts down rendered view by going back into shaded view and after that again attempts a viewport render - server crashes with following error: "ERROR! check console for detailed error message."
Various Errors from Console logs on numerous separate crashes while testing show this:
Octane: ERROR. Server log:
Failed devices:
NVIDIA GeForce RTX 5090(device 0), error message: "CUDA error (700) on device 0: an illegal memory access was encountered -> failed to allocate device memory"
Last Octane engine log entries:
CUDA error (700) on device 1: an illegal memory access was encountered
-> failed to load symbol data to the device(const24)
device 1: failed to upload light parameters for post.
Failed devices:
NVIDIA GeForce RTX 5090(device 0), error message: "CUDA error (700) on device 0: an illegal memory access was encountered -> could not get memory info"
Last Octane engine log entries:
OptiX error for device 1: [ERROR] Failed to copy internal launch buffer (CUDA error string: an illegal memory access was encountered, CUDA error code: 700)
Error recording resource event on user stream (CUDA error string: an illegal memory access was encountered, CUDA error code: 700)
OptiX API error for device 1: CUDA error
device 1: RTX tracing for post failed
CUDA error (700) on device 0: an illegal memory access was encountered
-> could not get memory info
CUDA error (700) on device 0: an illegal memory access was encountered
-> failed to allocate device memory
device 0: failed to allocate post-fx buffer
CUDA error (700) on device 0: an illegal memory access was encountered
-> failed to deallocate device memory
CUDA error (700) on device 0: an illegal memory access was encountered
-> could not get memory info
Strange thing is this issue only happens in viewport render. It doesn’t happen at all in camera render view or final render as long as viewport render hasn't been initiated (or sometimes if it was only initiated once).Failed devices:
NVIDIA GeForce RTX 5090(device 0), error message: "CUDA error (700) on device 0: an illegal memory access was encountered -> could not get memory info"
Last Octane engine log entries:
OptiX error for device 1: [ERROR] Failed to copy internal launch buffer (CUDA error string: an illegal memory access was encountered, CUDA error code: 700)
Error recording resource event on user stream (CUDA error string: an illegal memory access was encountered, CUDA error code: 700)
OptiX API error for device 1: CUDA error
device 1: RTX tracing for post failed
CUDA error (700) on device 1: an illegal memory access was encountered
-> failed to load symbol data to the device(const24)
device 1: failed to upload light parameters for post
CUDA error (700) on device 1: an illegal memory access was encountered
-> failed to load symbol data to the device(const24)
device 1: failed to upload light parameters for post
CUDA error (700) on device 0: an illegal memory access was encountered
-> could not get memory info
The error doesn’t occur on Mac. And I have only tried this with the current stable standard plugin not the add on beta. OctaneBlender version is 2025.3 - 30.7 [STABLE].
Can provide more detailed logs and specs of computers if necessary. Basically, I'm on Windows 10 using a 5090 and a 3090 on main machine and a 3080, 3080 Ti, and a 1080 Ti on the other.
NVIDIA Studio driver 581.29. Just updated to Studio driver 581.57 and issue persists (perhaps even more unstable).
The Mac, which tested fine, is a Mac Mini M1 running Sequoia 15.7.1.
Will post this in Discord channel as well.