Llama 2 7B › AMD Radeon AI PRO R9700
Llama 2 7B on the AMD Radeon AI PRO R9700
Best measured: 153 tokens/s generation and 5,908 tokens/s prompt processing (Q4_0, Vulkan, FA on).
Every run
| Quant | Backend | Generate tok/s | Prompt tok/s | Test | Build / by / notes | Source |
|---|---|---|---|---|---|---|
| Q4_0 | Vulkan FA on | 153 | 5,908 | tg128 | dd1ea52 | llama.cpp Vulkan scoreboard (#10879) |
| Q4_0 | Vulkan FA off | 146 | 5,610 | tg128 | dd1ea52 | llama.cpp Vulkan scoreboard (#10879) |
| Q4_0 | ROCm FA on | 98.0 | 4,773 | tg128 | bd4ef13 @gogich77 | llama.cpp ROCm scoreboard (#15021) |
| Q4_0 | ROCm FA off | 93.8 | 4,444 | tg128 | bd4ef13 @gogich77 | llama.cpp ROCm scoreboard (#15021) |
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.
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 Radeon AI PRO R9700 · Llama 2 7B on other hardware