Saved in this browser, used on every model page.
Yes
Yes. The Tesla P40 runs NLD-Uni fully in VRAM at Q8_0 (about 21 GB of 24 GB), at roughly 9.8 tokens/s.
- best quant
- Q8_0
- needs
- 21 GB
- usable
- 24 GB
- est. speed
- ~9.8 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the NLD-Uni page →
Every quantisation of NLD-Uni on the Tesla P40
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 7.4 GB | fits | ~32 tok/s |
| Q3_K_M | 10 GB | fits | ~21 tok/s |
| Q4_K_M | 12 GB | fits | ~17 tok/s |
| Q5_K_M | 14 GB | fits | ~15 tok/s |
| Q6_K | 16 GB | fits | ~13 tok/s |
| Q8_0 pick | 21 GB | fits | ~9.8 tok/s |
| FP16 / BF16 | 37 GB | offload | ~1.9 tok/s |
Speed is a bandwidth-based estimate for generating text (347 GB/s); prompt processing and long contexts are slower.
Measured speed on the Tesla P40
No community reports for this pair yet.
Report your own speed →Omni models that fit the Tesla P40 comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 17.5B | Q8_0 | ≈20 GBFits in memory | ~10 tok/s | |
| 12B | Q8_0 | ≈16 GBFits in memory | ~15 tok/s | |
| 24.3B | Q6_K | ≈21 GBFits in memory | ~9.5 tok/s | |
| 14.7B | Q8_0 | ≈17 GBFits in memory | ~12 tok/s | |
| 10.7B | Q8_0 | ≈8.9 GBFits in memory | ~17 tok/s | |
| 35.3BA3B | Q4_K_M | ≈19 GBFits in memory | ~105 tok/s |
Other GPUs that run NLD-Uni
Frequently asked questions
- Can the Tesla P40 run NLD-Uni?
- Yes. The Tesla P40 runs NLD-Uni fully in VRAM at Q8_0 (about 21 GB of 24 GB), at roughly 9.8 tokens/s.
- How fast does NLD-Uni run on the Tesla P40?
- An estimate from memory bandwidth gives about 9.8 tokens/s at Q8_0.
- Which quantisation of NLD-Uni should I use on the Tesla P40?
- Q8_0: the largest one that fits. Expected quality: lossless compared with the full-precision weights.
Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.