This repository contains my learnings and programs related to CUDA in C/C++ in no particular order. Any/every directory here will contain code(s) related to any topic. The codes will be heavily commented as well so anyone going through it would know what it is doing.
- softmax: Contains the kernels for naive, online, and optimized softmax operation.
- matmul: Contains the kernels for tiled matrix multiplication for 2D matrices.
- query-device: Contains the code for querying the device (GPU) properties.