Llama 3.1 8B Instruct › NVIDIA Tesla P100
Llama 3.1 8B Instruct on the NVIDIA Tesla P100
Best measured: 32.3 tokens/s generation and 399 tokens/s prompt processing (Q4_K_M, llamafile).
For comparison, our size-scaling estimate for this combination is 51.6 tok/s.
Every run
| Quant | Backend | Generate tok/s | Prompt tok/s | Test | Build / by / notes | Source |
|---|---|---|---|---|---|---|
| Q4_K_M | llamafile | 33.5 | 427 | LocalScore avg | LocalScore rank 120/422 | LocalScore (localscore.ai) |
| Q4_K_M | llamafile | 31.1 | 372 | LocalScore avg | LocalScore rank 129/422 | LocalScore (localscore.ai) |
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.
Rent one by the hour
Rent a GPU on RunPodAffiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.
All results for the NVIDIA Tesla P100 · Llama 3.1 8B Instruct on other hardware