🔌
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 H200 run LUXIA-21.4B-Alignment?
Yes — here's how fast and up to what context, with real measured numbers.
LUXIA-21.4B-Alignment on the NVIDIA H200
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 326k | 81.9-97.6 t/s | ~0.7-0.8 (slow) |
| Q8 | yes | 407k | 163.7-195.1 t/s | ~1.4-1.6 (slow) |
| Q4 | yes | 443k | 292.4-348.5 t/s | ~2.4-2.9 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to system RAM offload (depends on your system RAM) and speed drops to the "beyond" figure.
FAQ
Can the NVIDIA H200 run LUXIA-21.4B-Alignment?▶
Yes. At Q4 it generates ~292.4-348.5 tok/s and fits up to 443k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~2.4-2.9 tok/s.
How much context fits?▶
At Q4, up to about 443k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.