🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the NVIDIA DGX Spark run MatterSim (M3GNet - MatterSim-v1.0.0-5M)?
Yes — here's how fast and up to what context, with real measured numbers.
MatterSim (M3GNet - MatterSim-v1.0.0-5M) on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1735k | 22143.3-26390 t/s | ~811.1-966.7 (slow) |
| Q8 | yes | 1735k | 44286.7-52780 t/s | ~1622.2-1933.3 (slow) |
| Q4 | yes | 1735k | 79083.3-94250 t/s | ~2896.8-3452.4 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
→ All models on the NVIDIA DGX Spark
→ MatterSim (M3GNet - MatterSim-v1.0.0-5M) specs
→ Other hardware
FAQ
Can the NVIDIA DGX Spark run MatterSim (M3GNet - MatterSim-v1.0.0-5M)?▶
Yes. At Q4 it generates ~79083.3-94250 tok/s and fits up to 1735k context; beyond that it spills to SSD swap and slows to ~2896.8-3452.4 tok/s.
How much context fits?▶
At Q4, up to about 1735k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.