NVIDIA TensorRT 用 Global Performance Tuner 将性能优化扩展到单一构建路径之外（第一部分）
Scaling NVIDIA TensorRT Performance Optimization Beyond a Single Build Route with the Global Performance Tuner, Part 1
NVIDIA TensorRT 构建引擎时需在图层变换、融合、实现选择、内存布局和量化上做相互依赖的调度决策，单一构建路径难以覆盖全部最优解。Global Performance Tuner 将性能优化扩展到单一构建路径之外，第一部分对此展开说明。
Deploying a trained neural network involves more than translating each operator into a GPU kernel. It is a compilation problem with many legitimate answers. For...
Deploying a trained neural network involves more than translating each operator into a GPU kernel. It is a compilation problem with many legitimate answers. For example, when NVIDIA TensorRT builds an engine, it makes interdependent decisions about graph-IR transformations, fusion, implementation selection, memory placement, and quantization. This process is called scheduling.
来源：NVIDIA Developer Blog · developer.nvidia.com