AMD Instinct MI350X 288GBBest value
C49.7 tok/s decode
~$8,000 MSRP
Can it run?
Too heavy
Using Q4_K_M in vLLM
Fit status
Too heavy
Decode
11.0 tok/s
TTFT
17666 ms
Safe context
5K
Memory
310.3 GB / 96.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | F | Too heavy | 12.8 tok/s | 22025 ms | 10K |
| Chat | F | Too heavy | 12.8 tok/s | 8259 ms | 4K |
| Coding | F | Too heavy | 11.0 tok/s | 17666 ms | 5K |
| RAG | F | Too heavy | 12.8 tok/s | 27531 ms | 10K |
| Reasoning | F | Too heavy | 12.8 tok/s | 17895 ms | 5K |
How Qwen3-Coder 480B A35B Instruct (480B params) fits at each quantization level on RTX PRO 6000 Blackwell Server Edition 96GB (96.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 187.2 GB | Low | F0 |
Q3_K_S | 3 | 235.2 GB | Low | F0 |
NVFP4 | 4 |
Upgrade options
~$8,000 MSRP
268.8 GB |
| Medium |
| F0 |
Q4_K_M | 4 | 292.8 GB | Medium | F0 |
Q5_K_M | 5 | 345.6 GB | High | F0 |
Q6_K | 6 | 393.6 GB | High | F0 |
Q8_0 | 8 | 513.6 GB | Very High | F0 |
F16 | 16 | 984.0 GB | Maximum | F0 |