Llama 2 7B on the AMD FirePro D700

Best measured: 16.6 tokens/s generation and 70 tokens/s prompt processing (Q4_0, Vulkan, FA off).

Every run

Quant	Backend	Generate tok/s	Prompt tok/s	Test	Build / by / notes	Source
Q4_0	Vulkan FA off	16.6	70	tg128	d3bd719 MoltenVK, running in FP16 mode on FP32 only chip	llama.cpp Vulkan scoreboard (#10879)

tg128 = generating 128 tokens; pp512 = reading a 512-token prompt (llama-bench). LocalScore averages several prompt and output lengths, so its numbers run lower than tg128.

Check prices

Amazon Newegg

As an Amazon Associate I earn from qualifying purchases.

No card yet?

Try it on a rented GPU (RunPod)

Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.

All results for the AMD FirePro D700 · Llama 2 7B on other hardware