Server execution error: CUDA error.

[XILegacyy]

i use nvidia 5070 TI and this error message already trying all the thing online but nothing works help please

Server execution error: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

[RapFox-ink]

Same here...

"Server execution error: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions."

[XILegacyy]

its fine when im only using comfyui idk why

[arch0820]

Same error message after replacing with nvidia 5070.
I've been trying for 10 days without success. ㅠ
Should I take online payment for the time being?

Is this an issue that needs to be resolved by nvidiad?
Or is it an issue that needs to be resolved by KRITA AI?
Please help those involved.

Server execution error: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with TORCH_USE_CUDA_DSA to enable device-side assertions.

[XILegacyy]

i think the problem is the torch because torch is not updated to cuda 12.8?

[Acly]

You need to install ComfyUI with preview version of torch and cuda 12.8
https://blog.comfy.org/p/how-to-get-comfyui-running-on-your

#1586 #1620

[arch0820]

기다리던 답변 감사합니다. 다시 시도해보겠습니다. 2025년 4월 1일 (화) 오후 5:10, Acly ***@***.***>님이 작성:

…

You need to install ComfyUI with preview version of torch and cuda 12.8 https://blog.comfy.org/p/how-to-get-comfyui-running-on-your #1586 <#1586> #1620 <#1620> — Reply to this email directly, view it on GitHub <#1693 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BQ7WSHJF4ANJSXERWFT5KCL2XJCYFAVCNFSM6AAAAAB2C4HRM6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDONRYGU2DCMRTG4> . You are receiving this because you commented.Message ID: ***@***.***> [image: Acly]*Acly* left a comment (Acly/krita-ai-diffusion#1693) <#1693 (comment)> You need to install ComfyUI with preview version of torch and cuda 12.8 https://blog.comfy.org/p/how-to-get-comfyui-running-on-your #1586 <#1586> #1620 <#1620> — Reply to this email directly, view it on GitHub <#1693 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BQ7WSHJF4ANJSXERWFT5KCL2XJCYFAVCNFSM6AAAAAB2C4HRM6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDONRYGU2DCMRTG4> . You are receiving this because you commented.Message ID: ***@***.***>

[arch0820]

After replacing it with an Nvidia 5070, I get the same error message.
I've tried installing ComfyUI using the preview version of Torch and CUDA 12.8, but it won't run. I must be inexperienced. Is it unusable unless I fix it? Will it resolve itself if I wait? If it does resolve, how long will it take?

[arch0820]

Resolved There is another error
Error: Could not find LoRA 'lcm' (lcm-lora-sdv1-5.safetensors, lcm/sd1.5/pytorch_lora_weights.safetensors) used by sampler preset 'Realtime - LCM' Please advise