Can I run Nanbeige_Nanbeige4.3-33B-GGUF on a Radeon RX 9070 XT?
Yes, just
Its IQ2_M file (12.2 GB) fits in the 14.4 GB a Radeon RX 9070 XT can use, with room for ≈ 10K tokens of context. That file is under 3 bits per weight, so expect noticeably lower quality.
Full analysis of Nanbeige_Nanbeige4.3-33B-GGUF
Everything that runs on Radeon RX 9070 XT
Get the IQ2_M file on Hugging Face
ollama run hf.co/bartowski/Nanbeige_Nanbeige4.3-33B-GGUF:IQ2_MFiles and context on Radeon RX 9070 XT
Each file of Nanbeige_Nanbeige4.3-33B-GGUF (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.
| File | Size | On Radeon RX 9070 XT | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| BF16 | 66.2 GB | Doesn't fit | — | — |
| Q8_0 | 35.2 GB | Doesn't fit | — | — |
| Q6_K_L | 30.4 GB | Doesn't fit | — | — |
| Q6_K | 29.0 GB | Doesn't fit | — | — |
| Q6_K_S | 27.8 GB | Doesn't fit | — | — |
| Q5_K_M | 25.1 GB | Doesn't fit | — | — |
| Q5_K_S | 23.7 GB | Doesn't fit | — | — |
| Q4_K_L | 22.3 GB | Doesn't fit | — | — |
| Q4_1 | 21.2 GB | Doesn't fit | — | — |
| IQ4_NL | 20.9 GB | Doesn't fit | — | — |
| Q4_K_M | 20.8 GB | Doesn't fit | — | — |
| Q4_K_S | 19.7 GB | Doesn't fit | — | — |
| Q4_0 | 19.3 GB | Doesn't fit | — | — |
| IQ4_XS | 18.6 GB | Doesn't fit | — | — |
| IQ3_M | 17.8 GB | Doesn't fit | — | — |
| Q3_K_L | 17.3 GB | Doesn't fit | — | — |
| Q3_K_M | 15.9 GB | Doesn't fit | — | — |
| IQ3_XS | 15.2 GB | Doesn't fit | — | — |
| Q3_K_S | 15.2 GB | Doesn't fit | — | — |
| IQ3_XXS | 14.5 GB | Doesn't fit | — | — |
| Q2_K | 13.5 GB | Doesn't fit | 4K | 8K |
| IQ2_M | 12.2 GB | Tight | 10K | 20K |
| IQ2_S | 11.2 GB | Fits | 15K | 30K |
| IQ2_XS | 10.6 GB | Fits | 17K | 35K |
| IQ2_XXS | 10.1 GB | Fits | 20K | 40K |