selfhost.directory

The index of self-hostable projects — ranked, categorized and served with live updates.
Own your data and stop giving it to big tech.

tencent logo

Can the Tesla P40 run WeMM-Video-2B?

WeMM-Video-2B·3.3B·Tesla P40·24 GB VRAM

Saved in this browser, used on every model page.
Yes

Yes. The Tesla P40 runs WeMM-Video-2B fully in VRAM at FP16 / BF16 (about 7.3 GB of 24 GB), at roughly 29 tokens/s.

best quant
FP16 / BF16
needs
7.3 GB
usable
24 GB
est. speed
~29 tok/s

Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the WeMM-Video-2B page →

Every quantisation of WeMM-Video-2B on the Tesla P40

QuantQualityMemoryFitSpeed
Q2_K Heavy loss1.9 GBfits~177 tok/s
Q3_K_M Noticeable loss2.4 GBfits~119 tok/s
Q4_K_M Small loss2.8 GBfits~96 tok/s
Q5_K_M Near-lossless3.2 GBfits~82 tok/s
Q6_K Near-lossless3.5 GBfits~71 tok/s
Q8_0 Lossless4.3 GBfits~55 tok/s
FP16 / BF16 pickLossless7.3 GBfits~29 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 →

Vision models that fit the Tesla P40 comfortably

All that fit →
ModelParamsBest quantMemorySpeedLicence
4.3BQ8_0≈5.2 GBFits in memory~42 tok/sConditions
4.7BQ8_0≈5.9 GBFits in memory~39 tok/sCommercial OK
3.8BQ8_0≈4.0 GBFits in memory~48 tok/sNon-commercial
2.3BQ8_0≈2.8 GBFits in memory~79 tok/sCommercial OK
4.4BQ8_0≈5.8 GBFits in memory~40 tok/sCommercial OK
3.8BQ8_0≈6.9 GBFits in memory~47 tok/sCheck licence

Other GPUs that run WeMM-Video-2B

Frequently asked questions

Can the Tesla P40 run WeMM-Video-2B?
Yes. The Tesla P40 runs WeMM-Video-2B fully in VRAM at FP16 / BF16 (about 7.3 GB of 24 GB), at roughly 29 tokens/s.
How fast does WeMM-Video-2B run on the Tesla P40?
An estimate from memory bandwidth gives about 29 tokens/s at FP16 / BF16.
Which quantisation of WeMM-Video-2B should I use on the Tesla P40?
FP16 / BF16: 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.