Can the RTX 5090 run NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball?
NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball·33B·GeForce RTX 5090·32 GB VRAM
Yes. The RTX 5090 runs NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball fully in VRAM at Q6_K (about 28 GB of 32 GB), at roughly 36 tokens/s.
- best quant
- Q6_K
- needs
- 28 GB
- usable
- 32 GB
- est. speed
- ~36 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball page →
Every quantisation of NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball on the RTX 5090
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 12 GB | fits | ~91 tok/s |
| Q3_K_M | 17 GB | fits | ~61 tok/s |
| Q4_K_M | 20 GB | fits | ~49 tok/s |
| Q5_K_M | 24 GB | fits | ~42 tok/s |
| Q6_K pick | 28 GB | fits | ~36 tok/s |
| Q8_0 | 35 GB | offload | ~7.7 tok/s |
| FP16 / BF16 | 65 GB | too big | — |
Speed is a bandwidth-based estimate for generating text (1792 GB/s); prompt processing and long contexts are slower.
Measured speed on the RTX 5090
No community reports for this pair yet.
Report your own speed →Run NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball on the RTX 5090
vllm serve nvidia/NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketballservices:
vllm:
image: vllm/vllm-openai:latest
command: nvidia/NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball
ports:
- "8000:8000"
ipc: host
volumes:
- ~/.cache/huggingface:/root/.cache/huggingface
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: all
capabilities: [gpu]API on port 8000.
sglang serve nvidia/NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball --host 0.0.0.0 --port 30000services:
sglang:
image: lmsysorg/sglang:latest
command: sglang serve nvidia/NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball --host 0.0.0.0 --port 30000
ports:
- "30000:30000"
ipc: host
volumes:
- ~/.cache/huggingface:/root/.cache/huggingface
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: all
capabilities: [gpu]API on port 30000.
Omni models that fit the RTX 5090 comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 35.3BA3B | Q6_K | ≈30 GBFits in memory | ~399 tok/s | |
| 33BA3B | Q6_K | ≈28 GBFits in memory | ~399 tok/s | |
| 31.7BA3B | Q6_K | ≈27 GBFits in memory | ~399 tok/s | |
| 31.7BA3B | Q6_K | ≈27 GBFits in memory | ~399 tok/s | |
| 24.3B | Q8_0 | ≈26 GBFits in memory | ~38 tok/s | |
| 33.5B | Q6_K | ≈30 GBFits in memory | ~36 tok/s |
Other GPUs that run NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball
Frequently asked questions
- Can the RTX 5090 run NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball?
- Yes. The RTX 5090 runs NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball fully in VRAM at Q6_K (about 28 GB of 32 GB), at roughly 36 tokens/s.
- How fast does NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball run on the RTX 5090?
- An estimate from memory bandwidth gives about 36 tokens/s at Q6_K.
- Which quantisation of NVIDIA-NemotronLabs-AI-for-Media-Sports-Basketball should I use on the RTX 5090?
- Q6_K: the largest one that fits. Expected quality: near-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.