[GH-ISSUE #3276] Running Ollama with Zluda on AMD GPU for CUDA support #64056

Closed
opened 2026-05-03 16:00:10 -05:00 by GiteaMirror · 0 comments
Owner

Originally created by @drnushooz on GitHub (Mar 20, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/3276

What are you trying to do?

AMD has an official build of CUDA api on top of ROCm which is called Zluda. This is a placeholder of how ollama runs on various platform with AMD Radeon GPU. Here is the link to Zluda project https://github.com/vosen/ZLUDA

How should we solve this?

Just try to install Zluda on Linux and run Ollama and TensorFlow on it. See if it detects the GPU as CUDA compatible.

What is the impact of not solving this?

No response

Anything else?

No response

Originally created by @drnushooz on GitHub (Mar 20, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/3276 ### What are you trying to do? AMD has an official build of CUDA api on top of ROCm which is called Zluda. This is a placeholder of how ollama runs on various platform with AMD Radeon GPU. Here is the link to Zluda project https://github.com/vosen/ZLUDA ### How should we solve this? Just try to install Zluda on Linux and run Ollama and TensorFlow on it. See if it detects the GPU as CUDA compatible. ### What is the impact of not solving this? _No response_ ### Anything else? _No response_
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#64056