Advanced Profiling and Analytics for AMD Hardware

Omnitrace: Application Profiling, Tracing, and Analysis

AITemplate is a Python framework which renders neural network into high performance CUDA/HIP C++ code. Specialized for FP16 TensorCore (NVIDIA GPU) and MatrixCore (AMD GPU) inference.

Benchmarks to capture important workloads.

Convert nvprof profiles into about:tracing compatible JSON files