Massively Parallel Programming with CUDA

Explain the CUDA programming model: host, device, kernels, and thread hierarchy (threads, blocks, grids)

Manage GPU memory allocation and data transfers between host and device (cudaMalloc, cudaMemcpy)

Write and launch a basic CUDA kernel for an element-wise vector operation

Describe the different GPU memory spaces (global, shared, constant, registers) and their use cases

Optimize global memory access by ensuring memory coalescing

Use shared memory to reduce global memory traffic in a CUDA kernel

Synchronize threads within a block using __syncthreads() to coordinate computation

Implement a parallel reduction on the GPU using shared memory