🔌
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 AMD Instinct MI325X run TinyLlama-1.1B (1T token checkpoint)?
Yes — here's how fast and up to what context, with real measured numbers.
TinyLlama-1.1B (1T token checkpoint) on the AMD Instinct MI325X
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 3537k | 1718.2-2100 t/s | ~11.5-14 (slow) |
| Q8 | yes | 3554k | 3436.4-4200 t/s | ~22.9-28 (slow) |
| Q4 | yes | 3561k | 6136.4-7500 t/s | ~40.9-50 (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.
→ All models on the AMD Instinct MI325X
→ TinyLlama-1.1B (1T token checkpoint) specs
→ Other hardware
FAQ
Can the AMD Instinct MI325X run TinyLlama-1.1B (1T token checkpoint)?▶
Yes. At Q4 it generates ~6136.4-7500 tok/s and fits up to 3561k context; beyond that it spills to system RAM offload (depends on your system RAM) and slows to ~40.9-50 tok/s.
How much context fits?▶
At Q4, up to about 3561k tokens stay in fast memory; longer context spills to system RAM offload (depends on your system RAM) and throughput collapses.