optimization 6
- PDL 详解:让相邻 GPU kernel 首尾重叠——原理、写法、各框架实践与踩坑总结
- LLM 推理框架的 Autotune 设计指南:各家最佳实践与可落地的设计要点
- Piecewise CUDA Graph and Breakable CUDA Graph for LLM Inference
- When NVLS Is "Supported" on H100, Why Does AllReduce Take Another Path?
- Deep Dive: AllReduce and AllReduce Fusion in TensorRT-LLM
- Optimizing Nemotron v3 Super in TensorRT-LLM