Unable to get ComfyUI to install

[SunKrocket]

I've been trying to get AI Playground up and running on my system for a couple of days but I keep running into issues. Now It seems that ComfyUI won't install fully because PyTorch is not CUDA enabled. I don't have an Nvidia card, I'm trying to run off the CPU/iGPU. I'm pretty new to all this and I can't find anyone with this issue that DOESN'T have an Nvidia card. Any help is appreciated.

Win 11
Ryzen 7 8700g
32GB RAM -800+ GBs free

Could not autodetect AIMDO implementation, assuming Nvidia
Adding extra search path checkpoints C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\checkpoints
Adding extra search path loras C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\loras
Adding extra search path loras C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\lora
Adding extra search path vae C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\vae
Adding extra search path text_encoders C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\text_encoders
Adding extra search path text_encoders C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\clip
Adding extra search path clip_vision C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\clip_vision
Adding extra search path style_models C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\style_models
Adding extra search path embeddings C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\embeddings
Adding extra search path diffusers C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\diffusers
Adding extra search path vae_approx C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\vae_approx
Adding extra search path controlnet C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\controlnet
Adding extra search path controlnet C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\t2i_adapter
Adding extra search path gligen C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\gligen
Adding extra search path upscale_models C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\upscale_models
Adding extra search path upscale_models C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\latent_upscale_models
Adding extra search path hypernetworks C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\hypernetworks
Adding extra search path photomaker C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\photomaker
Adding extra search path classifiers C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\classifiers
Adding extra search path model_patches C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\model_patches
Adding extra search path audio_encoders C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\audio_encoders
Adding extra search path diffusion_models C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\diffusion_models
Adding extra search path diffusion_models C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\unet
Adding extra search path insightface C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\insightface
Adding extra search path facerestore_models C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\facerestore_models
Adding extra search path nsfw_detector C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\nsfw_detector
Adding extra search path inpaint C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\models\ComfyUI\inpaint
Setting output directory to: C:\Users\M. Michaux\Documents\AI-Playground\media
[ComfyUI-Manager] python -m pip not available. Falling back to uv.
[ComfyUI-Manager] Using uv as Python module for pip operations.

Prestartup times for custom nodes:
5.0 seconds: C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI\custom_nodes\ComfyUI-Manager

C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI.venv\Lib\site-packages\torch\xpu_init_.py:61: UserWarning: XPU device count is zero! (Triggered internally at C:\actions-runner_work\pytorch\pytorch\pytorch\c10\xpu\XPUFunctions.cpp:115.)
return torch._C.xpu_getDeviceCount()
Found comfy_kitchen backend triton: {'available': False, 'disabled': True, 'unavailable_reason': 'Neither CUDA nor XPU available on this system', 'capabilities': []}
Found comfy_kitchen backend cuda: {'available': False, 'disabled': True, 'unavailable_reason': 'CUDA not available on this system', 'capabilities': []}
Found comfy_kitchen backend eager: {'available': True, 'disabled': False, 'unavailable_reason': None, 'capabilities': ['apply_rope', 'apply_rope1', 'dequantize_mxfp8', 'dequantize_nvfp4', 'dequantize_per_tensor_fp8', 'quantize_mxfp8', 'quantize_nvfp4', 'quantize_per_tensor_fp8', 'scaled_mm_mxfp8', 'scaled_mm_nvfp4']}
Checkpoint files will always be loaded safely.
Traceback (most recent call last):
File "C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI\main.py", line 197, in
import execution
File "C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI\execution.py", line 17, in
import comfy.model_management
File "C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI\comfy\model_management.py", line 256, in
total_vram = get_total_memory(get_torch_device()) / (1024 * 1024)
^^^^^^^^^^^^^^^^^^
File "C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI\comfy\model_management.py", line 206, in get_torch_device
return torch.device(torch.cuda.current_device())
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI.venv\Lib\site-packages\torch\cuda_init.py", line 1069, in current_device
lazy_init()
File "C:\Users\M. Michaux\AppData\Local\Programs\AI Playground\resources\ComfyUI.venv\Lib\site-packages\torch\cuda_init.py", line 403, in _lazy_init
raise AssertionError("Torch not compiled with CUDA enabled")
AssertionError: Torch not compiled with CUDA enabled

[bobduffy]

Use the recent hotfix
https://github.com/intel/AI-Playground/releases/tag/v3.1.0-alpha_hf

[SunKrocket]

Thanks bobduffy. I was able to get OpenVino and Llama going using the hotfix but I'm still seeing the same error and I can't start ComfyUI. I manually deleted everything in the temp folder because it seems like the installation is reverting back to settings from my initial install. I also ran a registry scan and I deleted the obsolete entries. I thought the hotfix was the solution. I even deleted the file I download the other day and reinstalled from a fresh download. I asked the AI about my issue and it told me that I need to install a version of pytorch that supports Nvidia CUDA. Not sur how to do that, but is that viable?

[bobduffy]

I now see you're running on Ryzen. AI Playground's installation supports PyTorch for Intel xpu (Arc GPUs)or Nvidia CUDA. We do not currently have an installer script to install for AMDs ROCm.

Unfortunately the way PyTorch works is you need to define the installation scripts specific to a manufacturer's GPU

You might be able to get AI to tell you how to change the AI Playground installer to ROCm or do a post installation CLI to the ComfyUI environment to install for torch ROCm