can-i-run-this-llm

Can I run this LLM?

Can a AMD Instinct MI355X (288GB) run Hunyuan Hy3 (MoE)?

Hunyuan · Apache 2.0 · estimated for a 288.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Hunyuan Hy3 (MoE) at Q5_K_M · 212.8 GB · 182–273 tok/s on a AMD Instinct MI355X (288GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K106.7 GB237–355 tok/sRuns great
Q3_K_M137.6 GB218–327 tok/sRuns great
Q4_K_M182.2 GB195–293 tok/sRuns great
Q5_K_M212.8 GB182–273 tok/sRuns great
Q6_K257.3 GB166–249 tok/sWon't fit
Q8_0317.7 GB148–223 tok/sWon't fit

Ran Hunyuan Hy3 (MoE) on a AMD Instinct MI355X (288GB)? Add your real numbers →