🔌
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 Jetson T3000 run minChatGPT?
Yes — here's how fast and up to what context, with real measured numbers.
minChatGPT on the NVIDIA Jetson T3000
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 367k | 60.8-72.5 t/s | ~2.2-2.7 (slow) |
| Q8 | yes | 392k | 121.6-144.9 t/s | ~4.5-5.3 (slow) |
| Q4 | yes | 403k | 217.1-258.8 t/s | ~8-9.5 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the NVIDIA Jetson T3000 run minChatGPT?▶
Yes. At Q4 it generates ~217.1-258.8 tok/s and fits up to 403k context; beyond that it spills to SSD swap and slows to ~8-9.5 tok/s.
How much context fits?▶
At Q4, up to about 403k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.