Model Anatomy See what's inside open-weight AI models

Can I run Hy-Memory-30B-A3B on an RTX 3060 12 GB?

Yes, with a GGUF version

The published weights are too big, but the Hy-Memory-30B-A3B-Q8_0-expsIQ2_XXS file from tencent/Hy-Memory-30B-A3B-GGUF (9.14 GB) fits, with room for ≈ 16K tokens of context.

Full analysis of Hy-Memory-30B-A3B Everything that runs on RTX 3060 12 GB Get the Hy-Memory-30B-A3B-Q8_0-expsIQ2_XXS file on Hugging Face
ollama run hf.co/tencent/Hy-Memory-30B-A3B-GGUF

Files and context on RTX 3060 12 GB

Each file of Hy-Memory-30B-A3B (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

FileSizeOn RTX 3060 12 GB Context (16-bit)Context (8-bit)
As published60.1 GBDoesn't fit——
Hy-Memory-30B-A3B-Q8_0tencent/Hy-Memory-30B-A3B-GGUF32.0 GBDoesn't fit——
Q6_Ktencent/Hy-Memory-30B-A3B-GGUF24.7 GBDoesn't fit——
Q4_K_Mtencent/Hy-Memory-30B-A3B-GGUF18.2 GBDoesn't fit——
Q3_K_Mtencent/Hy-Memory-30B-A3B-GGUF14.4 GBDoesn't fit——
Hy-Memory-30B-A3B-Q8_0-expsIQ2_XXStencent/Hy-Memory-30B-A3B-GGUF9.14 GBTight16K33K
Q8_0tencent/Hy-Memory-30B-A3B-GGUF8.03 GBFits27K55K

Hy-Memory-30B-A3B on other hardware

Popular models that run on RTX 3060 12 GB

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.