New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
Can the Instinct MI455X run BERT-RBP?
Yes — here's how fast and up to what context, with real measured numbers.
BERT-RBP on the Instinct MI455X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 6038k | 56127.3-68600 t/s | ~114.5-140 (slow) |
| Q8 | yes | 6040k | 112254.5-137200 t/s | ~229.1-280 (slow) |
| Q4 | yes | 6041k | 200454.5-245000 t/s | ~409.1-500 (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 Instinct MI455X run BERT-RBP?▶
Yes. At Q4 it generates ~200454.5-245000 tok/s and fits up to 6041k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~409.1-500 tok/s.
How much context fits?▶
At Q4, up to about 6041k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.