dnv2003 / v100-skinny
Hand-written NVFP4 W4A16 CUDA kernels for Volta
Why this job: topics: llm-inference, vllm, inference-engine, llm-serving, quantization
What it does
This is a same-lab system comparison on the same Qwen3.8 base model, not a same-weight causal engine A/B: the two engines use different published NVFP4 artifacts. NInfer depth 5 is its maximum exposed depth and its best measured setting on this fixture; v100-skinny's best measured setting is k=7.
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit 2 months ago
- Stars
- 182
- +2 this week
- Licence
- Custom
- Read before business use
- Downloads for
- None listed
- No release files
- Latest version
- None yet
- No releases
- Built with
- Python
- 90% of the code
Is it looked after?
- Contributors
- 1
- Last 90 days
- Issues
- 0% closed
- 4 still open
- Pull requests
- 0 merged
- 4 waiting
- Releases
- None
- Never published
- No contributing guide
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| pegainfer-project / pegainfer | 739 | yesterday | Apache-2.0 | Linux |
| SemiAnalysisAI / InferenceX | 2,021 | today | Apache-2.0 | No downloads |
| liguodongiot / llm-action | 25,271 | 28 days ago | Apache-2.0 | No downloads |
More: languages, topics and project record
Languages
- Python 90%
- Cuda 7%
- Shell 2%
Topics
- cuda
- cuda-kernels
- fp4
- gemm
- gpu
- inference-engine
- int4
- llm-inference
- llm-serving
- nvfp4
- nvidia
- performance-optimization
Project record
- JobLocal LLMs
- StatusListed
- Commits7
- Forks26
- Watching5
- Stars in 28 days+15
- CreatedAug 11, 2026
- First seen by usNov 7, 2026
- CheckedNov 11, 2026