This repository contains my solutions to the CUDA laboratories of the course Parallel Computing with GPUs and some notes on the course content.
The provided solutions contain up-to-date versions (as of June 2023) of deprecated primitives and functionalities referred to in the labs descriptions, e.g. we use cudaOccupancyMaxActiveBlocksPerMultiprocessor instead of the deprecated CUDA Occupancy Calculator.
We also use the profiling tool Nsight Compute Cli. Notice we use the light-weight (not-interactive) Nsight Compute Cli and not Nsight Compute.