Originally published at: https://developer.nvidia.com/blog/boosting-productivity-and-performance-with-the-nvidia-cuda-11-2-c-compiler/
The 11.2 CUDA C++ compiler incorporates features and enhancements aimed at improving developer productivity and the performance of GPU-accelerated applications. The compiler toolchain gets an LLVM upgrade to 7.0, which enables new features and can help improve compiler code generation for NVIDIA GPUs. Link-time optimization (LTO) for device code (also known as device LTO), introduced…