Llama 2 7B › AMD Radeon RX 6900 XT
Llama 2 7B on the AMD Radeon RX 6900 XT
Best measured: 108 tokens/s generation and 1,901 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 | 108 | 1,901 | tg128 | a972fae | llama.cpp Vulkan scoreboard (#10879) |
| Q4_0 | Vulkan FA on | 106 | 1,762 | tg128 | a972fae | llama.cpp Vulkan scoreboard (#10879) |
| Q4_0 | ROCm FA off | 88.5 | 1,890 | tg128 | a972fae @notgood | llama.cpp ROCm scoreboard (#15021) |
| Q4_0 | ROCm FA on | 85.0 | 1,948 | tg128 | a972fae @notgood | 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 RX 6900 XT · Llama 2 7B on other hardware