Can it run?
Too heavy
Using Q4_K_M in vLLM
Fit status
Too heavy
Decode
13.0 tok/s
TTFT
14846 ms
Safe context
4K
Memory
630.2 GB / 128.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | F | Too heavy | 13.0 tok/s | 21595 ms | 6K |
| Chat | F | Too heavy | 13.0 tok/s | 8098 ms | 4K |
| Coding | F | Too heavy | 13.0 tok/s | 14846 ms | 4K |
| RAG | F | Too heavy | 13.0 tok/s | 26993 ms | 6K |
| Reasoning | F | Too heavy | 13.0 tok/s | 17546 ms | 4K |
How Kimi K2.5 (1000B params) fits at each quantization level on Gaudi 3 128GB (128.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 390.0 GB | Low | F0 |
Q3_K_S | 3 | 490.0 GB | Low | F0 |
NVFP4 | 4 | 560.0 GB | Medium | F0 |
Q4_K_M | 4 | 610.0 GB | Medium | F0 |
Q5_K_M | 5 | 720.0 GB | High | F0 |
Q6_K | 6 | 820.0 GB | High | F0 |
Q8_0 | 8 | 1070.0 GB | Very High | F0 |
F16 | 16 | 2050.0 GB | Maximum | F0 |
