 |
VOOZH | about |
MPT-30B-Instruct needs ~49.4 GB but AMD Instinct MI100 32GB only has 32.0 GB. Try a smaller quantization or lighter model.
Operating mode
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
17.4 GB over capacity — needs offload or smaller quantization
Fit status
Too heavy
Decode
11.3 tok/s
TTFT
17109 ms
Safe context
4K
Memory
49.4 GB / 32.0 GB
Offload
40%
Usable VRAM is the main blocker for this model.
Not enough usable memory
The model needs 49.4 GB, but this setup only exposes 32.0 GB of usable VRAM.
Add more VRAM headroom
The first useful upgrade is more dedicated VRAM so you can fit the model without shrinking context or dropping to a much lower quant.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | C | Very compromised (needs ~3.3 GB host RAM) | 20.0 tok/s | 5280 ms | 4K |
| Coding | F | Too heavy | 11.3 tok/s | 17109 ms | 4K |
| Agentic Coding | F | Too heavy | 5.7 tok/s | 49806 ms | 4K |
| Reasoning | F | Too heavy | 11.3 tok/s | 20220 ms | 4K |
| RAG | F | Too heavy | 5.7 tok/s | 62258 ms | 4K |
How MPT-30B-Instruct (30B params) fits at each quantization level on AMD Instinct MI100 32GB (32.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 11.7 GB | Low | B68 |
Q3_K_S | 3 | 14.7 GB | Low | B69 |
NVFP4 | 4 |
Upgrade options
Makes the model fit on the accelerator instead of staying completely out of reach.
Raises estimated decode speed by about 41%.
~$3,999 MSRP
Makes the model fit on the accelerator instead of staying completely out of reach.
Raises estimated decode speed by about 41%.
~$3,999 MSRP
Makes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$8,000 MSRP
Makes the model fit on the accelerator instead of staying completely out of reach.
Removes host-memory offload, which is usually the single biggest latency and throughput win.
~$40,000 MSRP
16.8 GB |
| Medium |
| A70 |
Q4_K_M | 4 | 18.3 GB | Medium | B70 |
Q5_K_M | 5 | 21.6 GB | High | B70 |
Q6_KBest for your GPU | 6 | 24.6 GB | High | B69 |
Q8_0 | 8 | 32.1 GB | Very High | F0 |
F16 | 16 | 61.5 GB | Maximum | F0 |
Add more VRAM headroom. The first useful upgrade is more dedicated VRAM so you can fit the model without shrinking context or dropping to a much lower quant.
