M3.1 Flash on a single 256GB M5 Ultra via Metal runs poorly, poster says

The poster ran M3.1 Flash (428B-A23B) on one 256GB M5 Ultra with TensorFold's native Metal engine, judged it not working well, guessed two DGX Sparks would do better, and shared an MLX repo.

𝙲𝚘𝚛𝚗𝚘𝚔𝚞𝚗

@cornokunArchived 3 days agoOriginally posted 2026.10.24

𝙲𝚘𝚛𝚗𝚘𝚔𝚞𝚗 @cornokunOriginally posted 2026.10.24

Open original on X

I cooked something again. Doesn't work very well on Metal in 256gb IMO. Imagine it would run better across two DGX Sparks. MiniMax-M3.1-Flash (428B-A23B) on one 256 GB M5 Ultra with TensorFold's native Metal engine https://github.com/BobClawblaw/minimax-m31-flash-mlx-m5-ultra

x.com/cornokun/status/2113958551681114228Archived 2026.10.25 12:07

Open original on X