Learning to write cuda kernels with 100 Days of CUDA challenge
Summary:
Implemented vector addition by writing a simple CUDA program. Explored how to launch a kernel to perform a parallelized addition of two arrays, where each thread computes the sum of a pair of values.
Learned:
- Basics of writing a CUDA kernel.
- Understanding of grid, block, and thread hierarchy in CUDA.
- How to allocate and manage device (GPU) memory using
cudaMalloc,cudaMemcpy, andcudaFree.
- Read Chapter 1 of the PMPP book.
- Learned about the fundamentals of parallel programming, CUDA architecture, and the GPU execution model.