Can I run Kimi-Linear-48B-A3B-Base on an RTX 6000 Ada?
Probably, at 4-bit
The published weights are too big, but a 4-bit version (about 27.6 GB) would fit, with room for ≈ 2.1M tokens of context. The site hasn't analyzed a GGUF of Kimi-Linear-48B-A3B-Base yet.
Files and context on RTX 6000 Ada
Each file of Kimi-Linear-48B-A3B-Base (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 RTX 6000 Ada | Context (16-bit) | Context (8-bit) |
|---|---|---|---|---|
| As published | 98.2 GB | Doesn't fit | — | — |
| 4-bit (estimate) | 27.6 GB | Fits | 2.1M | 4.1M |