Llama 2 7B › NVIDIA RTX 5060 Ti
Llama 2 7B on the NVIDIA RTX 5060 Ti
Best measured: 97.0 tokens/s generation and 3,913 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 | 97.0 | 3,913 | tg128 | 89f10ba coopmat2 | llama.cpp Vulkan scoreboard (#10879) |
| Q4_0 | Vulkan FA off | 93.5 | 3,461 | tg128 | 89f10ba coopmat2 | llama.cpp Vulkan scoreboard (#10879) |
| Q4_0 | CUDA FA on | 93.5 | 4,196 | tg128 | 89d1029 @mike-llamacpp | llama.cpp CUDA scoreboard (#15013) |
| Q4_0 | CUDA FA off | 90.9 | 3,737 | tg128 | 89d1029 @mike-llamacpp | llama.cpp CUDA scoreboard (#15013) |
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 NVIDIA RTX 5060 Ti · Llama 2 7B on other hardware