From 4fa0254f0b0d84e6ec0f1cf074822e812235df4d Mon Sep 17 00:00:00 2001 From: Francisco Herrera Date: Sat, 13 Dec 2025 23:48:20 -0500 Subject: [PATCH 1/3] Add message for pre-RDNA AMD GPU support via opencl --- docs/backend/OPENCL.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/backend/OPENCL.md b/docs/backend/OPENCL.md index e52baffdff..260caf52d1 100644 --- a/docs/backend/OPENCL.md +++ b/docs/backend/OPENCL.md @@ -17,7 +17,7 @@ OpenCL (Open Computing Language) is an open, royalty-free standard for cross-pla ### Llama.cpp + OpenCL -The llama.cpp OpenCL backend is designed to enable llama.cpp on **Qualcomm Adreno GPU** firstly via OpenCL. Thanks to the portabilty of OpenCL, the OpenCL backend can also run on certain Intel GPUs although the performance is not optimal. +The llama.cpp OpenCL backend is designed to enable llama.cpp on **Qualcomm Adreno GPU** firstly via OpenCL. Thanks to the portabilty of OpenCL, the OpenCL backend can also run on certain Intel GPUs although the performance is not optimal, as well as pre-RDNA AMD GPUs which do not support HIP. ## OS From 9fa94920c07f23e2bc228eea297f3d2f299dfcdf Mon Sep 17 00:00:00 2001 From: Francisco Herrera Date: Sun, 14 Dec 2025 00:14:21 -0500 Subject: [PATCH 2/3] Update OpenCL.md with AMD GPU exceptions Added a note about exceptions for AMD GPUs in the OpenCL backend section. --- docs/backend/OPENCL.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/backend/OPENCL.md b/docs/backend/OPENCL.md index 260caf52d1..d73265b178 100644 --- a/docs/backend/OPENCL.md +++ b/docs/backend/OPENCL.md @@ -17,7 +17,7 @@ OpenCL (Open Computing Language) is an open, royalty-free standard for cross-pla ### Llama.cpp + OpenCL -The llama.cpp OpenCL backend is designed to enable llama.cpp on **Qualcomm Adreno GPU** firstly via OpenCL. Thanks to the portabilty of OpenCL, the OpenCL backend can also run on certain Intel GPUs although the performance is not optimal, as well as pre-RDNA AMD GPUs which do not support HIP. +The llama.cpp OpenCL backend is designed to enable llama.cpp on **Qualcomm Adreno GPU** firstly via OpenCL. Thanks to the portabilty of OpenCL, the OpenCL backend can also run on certain Intel GPUs although the performance is not optimal, as well as pre-RDNA AMD GPUs which do not support HIP, with [few exceptions.](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/system-requirements.html) ## OS From 90a76f56dfebc92df7b7ed17429baa4c87fbfc3a Mon Sep 17 00:00:00 2001 From: Francisco Herrera Date: Sun, 14 Dec 2025 00:29:34 -0500 Subject: [PATCH 3/3] clarify meaning --- docs/backend/OPENCL.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/backend/OPENCL.md b/docs/backend/OPENCL.md index d73265b178..36278654d6 100644 --- a/docs/backend/OPENCL.md +++ b/docs/backend/OPENCL.md @@ -17,7 +17,7 @@ OpenCL (Open Computing Language) is an open, royalty-free standard for cross-pla ### Llama.cpp + OpenCL -The llama.cpp OpenCL backend is designed to enable llama.cpp on **Qualcomm Adreno GPU** firstly via OpenCL. Thanks to the portabilty of OpenCL, the OpenCL backend can also run on certain Intel GPUs although the performance is not optimal, as well as pre-RDNA AMD GPUs which do not support HIP, with [few exceptions.](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/system-requirements.html) +The llama.cpp OpenCL backend is designed to enable llama.cpp on **Qualcomm Adreno GPU** firstly via OpenCL. Thanks to the portabilty of OpenCL, the OpenCL backend can also run on certain Intel GPUs although the performance is not optimal, as well as pre-RDNA AMD GPUs, which do not support HIP with [few exceptions.](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/system-requirements.html) ## OS