internlm2 math plus 20b i1 needs ~28.2 GB VRAM. Gaudi 3 128GB has 128.0 GB. With Q4_K_M quantization, expect ~212 tok/s.
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
Fit status
Runs well
Decode
212.3 tok/s
TTFT
912 ms
Safe context
697K
Memory
28.2 GB / 128.0 GB
The raw memory story may look fine, but the software ecosystem is still a constraint here.
Runtime ecosystem is narrower than CUDA
Intel GPUs can look attractive on memory per dollar, but local AI tooling, kernels, and model coverage are still broader and easier on CUDA today.
Prefer CUDA if you want the path of least resistance
If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | C | Runs well | 212.3 tok/s | 497 ms | 697K |
| Coding | C | Runs well | 212.3 tok/s | 912 ms | 697K |
| Agentic Coding | C | Runs well | 212.3 tok/s | 1326 ms | 697K |
| Reasoning | C | Runs well | 212.3 tok/s | 1078 ms | 697K |
| RAG | C | Runs well | 212.3 tok/s | 1658 ms | 697K |
How internlm2 math plus 20b i1 (20B params) fits at each quantization level on Gaudi 3 128GB (128.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 7.8 GB | Low | D38 |
Q3_K_S | 3 | 9.8 GB | Low | D38 |
NVFP4 | 4 |
Copy-paste commands to run internlm2 math plus 20b i1 on your machine.
Run
lms load hf-mradermacher--internlm2-math-plus-20b-i1-gguf && lms server startYes, Gaudi 3 128GB can run internlm2 math plus 20b i1 with a C grade (Runs well). Expected decode speed: 212.3 tok/s.
internlm2 math plus 20b i1 (20B parameters) requires approximately 28.2 GB of memory with Q4_K_M quantization.
The recommended quantization for internlm2 math plus 20b i1 is Q4_K_M, which balances quality and memory efficiency.
On Gaudi 3 128GB, internlm2 math plus 20b i1 achieves approximately 212.3 tokens per second decode speed with a time-to-first-token of 912ms using Q4_K_M quantization.
For coding workloads, internlm2 math plus 20b i1 on Gaudi 3 128GB receives a C grade with 212.3 tok/s and 697K context.
On Gaudi 3 128GB, internlm2 math plus 20b i1 can safely use up to 697K tokens of context. The model's official context limit is —, but available memory constrains the safe maximum.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/hf-mradermacher--internlm2-math-plus-20b-i1-gguf-on-gaudi-3-128gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview:
11.2 GB |
| Medium |
| D38 |
Q4_K_M | 4 | 12.2 GB | Medium | D38 |
Q5_K_M | 5 | 14.4 GB | High | D38 |
Q6_K | 6 | 16.4 GB | High | D38 |
Q8_0 | 8 | 21.4 GB | Very High | D39 |
F16Best for your GPU | 16 | 41.0 GB | Maximum | C42 |
Prefer CUDA if you want the path of least resistance. If your goal is maximum runtime coverage, easier troubleshooting, and better support for new local AI releases, CUDA is usually still the safer upgrade path.
Often yes, if your goal is the easiest setup and the widest runtime support. Intel can offer attractive memory capacity, but CUDA still tends to win on tooling maturity, guides, kernels, and model coverage for local AI.