This project demonstrates the integration of a CUDA kernel within an NVIDIA Holoscan application. It consists of two custom operators: one for memory allocation and data initialization, and another for executing the CUDA kernel. The application was profiled using Nsight systems and the kernel with Nsight compute

Central repository for applications and operators for Holoscan