Gemma 4-31B / Vulkan RADV native PP512/TG128

on AMD Strix Halo 128GB · llama.cpp / Vulkan RADV / build unknown

Oct 13, 2026

Throughput

11.1 t/s gen · 261.0 tokens/s prompt processing (prefill)Prompt processing measures input prefill speed; it does not tell you the time to first token.

Quant

UD-Q4_K_XL (GGUF)

Flash Attention

System RAM

128 GB

Summary

User reports Gemma-4-31B at 11.1 t/s generation and 261 t/s prompt processing on an AMD Strix Halo 128GB system with a single Radeon 8060S iGPU. Setup is llama.cpp via RADV with the Unsloth UD-Q4_K_XL quant, full offload, batch 2048 and flash attention. The 261 t/s is native PP512 and the 11.1 t/s is native TG128, separate microbenchmarks rather than one end-to-end request; the same 11.1 t/s generation is reported for AMDVLK.