[AutoDeploy] BTK findings: Nano high concurrencies
Open
Nobody has claimed this yet.
feature request
General perf
- Dominant language
- Python
- Stars
- 14.7k
- Forks
- 2.8k
- Avg merge
- 2d 23h
- Merged PRs (30d)
- 489
Description
🚀 The feature, motivation and pitch
Nano FP8 TRT-LLM (AD and PT backends) saturates at conc=128 (similarly for NVFP4 at conc=64).
vLLM is able to scale further conc=256,512 and possibly beyond that (current sweep stops at 512).
For lower concurrencies, all three frameworks (TRT-LLM AD, TRT-LLM PT and vLLM) are on-par.
Capture traces, root cause the reason for the perf gap and fix.
Alternatives
No response
Additional context
No response
Before submitting a new issue...
- Make sure you already searched for relevant issues, and checked the documentation and examples for answers to frequently asked questions.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.