🔌
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 bge-reranker-large?
Yes — here's how fast and up to what context, with real measured numbers.
bge-reranker-large on the NVIDIA DGX Spark
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1718k | 177.9-212.1 t/s | ~6.5-7.8 (slow) |
| Q8 | yes | 1726k | 355.9-424.1 t/s | ~13-15.5 (slow) |
| Q4 | yes | 1730k | 635.5-757.4 t/s | ~23.3-27.7 (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 DGX Spark run bge-reranker-large?▶
Yes. At Q4 it generates ~635.5-757.4 tok/s and fits up to 1730k context; beyond that it spills to SSD swap and slows to ~23.3-27.7 tok/s.
How much context fits?▶
At Q4, up to about 1730k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.