🔌
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 Jetson Thor T3000 run TAPE Transformer?
Yes — here's how fast and up to what context, with real measured numbers.
TAPE Transformer on the Jetson Thor T3000
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 415k | 2622.2-3125.1 t/s | ~96.1-114.5 (slow) |
| Q8 | yes | 416k | 5244.5-6250.3 t/s | ~192.1-228.9 (slow) |
| Q4 | yes | 416k | 9365.1-11161.2 t/s | ~343-408.8 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the Jetson Thor T3000 run TAPE Transformer?▶
Yes. At Q4 it generates ~9365.1-11161.2 tok/s and fits up to 416k context; beyond that it spills to SSD swap and slows to ~343-408.8 tok/s.
How much context fits?▶
At Q4, up to about 416k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.