llama.cpp Vulkan Radeon Pro W5700 llama-2-7b Q4_0 pp512 FA
446.98 tok/s (pp512)
Fields
id
benchmarks/llamacpp-vulkan-radeon-pro-w5700-llama27b-q40-fa-pp512Metric
prefill_tok_s
Value
446.98
Unit
tok/s (pp512)
Model
llama-2-7b.Q4_0.gguf (ggml-org, 3.56 GiB, 6.74B params)
Methodology
COMMUNITY-SCOREBOARD (llama.cpp discussion #10879, maintainer-curated Vulkan backend). llama-bench llama-2-7b.Q4_0.gguf (ggml-org, 3.56 GiB, 6.74B params), pp512 and tg128, batch 1, full GPU offload, Vulkan. Per-request batch-1, not aggregate across concurrent requests. Flash attention on (-fa 1). Build 23bc779. Chip AMD Radeon Pro W5700. Printed mean 446.98 +/- 0.39.
Measured by
third_party
Reproducible
yes
notes
Flash-attention scoreboard row. Prefill is a wash versus noFA (446.98 vs 449.85).
Notes
Flash-attention scoreboard row. Prefill is a wash versus noFA (446.98 vs 449.85).