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 Segatron XL large, M=384?
Yes — here's how fast and up to what context, with real measured numbers.
Segatron XL large, M=384 on the Instinct MI455X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 6034k | 24023.3-29361.9 t/s | ~49-59.9 (slow) |
| Q8 | yes | 6038k | 48046.7-58723.7 t/s | ~98.1-119.8 (slow) |
| Q4 | yes | 6039k | 85797.7-104863.8 t/s | ~175.1-214 (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 Segatron XL large, M=384?▶
Yes. At Q4 it generates ~85797.7-104863.8 tok/s and fits up to 6039k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~175.1-214 tok/s.
How much context fits?▶
At Q4, up to about 6039k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.