AMD ROCm 7.1 vs. RADV Vulkan For Llama.cpp With The Radeon AI PRO R9700

Written by Michael Larabel in Display Drivers on 18 November 2025 at 10:45 AM EST. Page 2 of 4. 24 Comments.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: Qwen3-8B-Q8_0, Test: Text Generation 128). Vulkan was the fastest.

Text generation performance with Llama.cpp was one of the first areas where the Vulkan back-end was performing better than ROCm. Even with the new ROCm 7.1 release, it's still that way with this RDNA4 GPU performing better on the RADV driver with Vulkan back-end.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: Qwen3-8B-Q8_0, Test: Prompt Processing 512). AMD ROCm HIP was the fastest.

Prompt processing performance is more mixed between ROCm and Vulkan depending upon the particular large language model and other factors.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: Qwen3-8B-Q8_0, Test: Prompt Processing 1024). AMD ROCm HIP was the fastest.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: Qwen3-8B-Q8_0, Test: Prompt Processing 2048). AMD ROCm HIP was the fastest.

With Qwen3 8B the ROCm back-end provided better prompt processing performance on the Radeon AI PRO R9700 graphics card.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: gpt-oss-20b-Q8_0, Test: Text Generation 128). Vulkan was the fastest.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: gpt-oss-20b-Q8_0, Test: Prompt Processing 512). Vulkan was the fastest.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: gpt-oss-20b-Q8_0, Test: Prompt Processing 1024). Vulkan was the fastest.

👁 Llama.cpp benchmark with settings of Backend Comparison (Model: gpt-oss-20b-Q8_0, Test: Prompt Processing 2048). Vulkan was the fastest.

With GPT-OSS 20B, the Vulkan back-end was faster for both text generation and prompt processing on this AMD graphics card.

24 Comments - Next Page

Page: 1 2 3 4 Next Page

URL: https://www.phoronix.com/review/rocm-71-llama-cpp-vulkan/2

⇱ AMD ROCm 7.1 vs. RADV Vulkan For Llama.cpp With The Radeon AI PRO R9700 - Phoronix

AMD ROCm 7.1 vs. RADV Vulkan For Llama.cpp With The Radeon AI PRO R9700