Worked examples

Complete programs that combine the pieces described in the guides.

• A portable diffusion solver: array-level code only, one script for CPU and GPU, with a command-line backend switch, timing and output at the boundaries.
• Porting a particle-in-cell code: a small simulation whose kernels are ported to CUDA one at a time with a KernelCatalog, verified with parity tests, and checked for transfers.

For an MPI program with one rank per GPU, see the start-up sequence and halo exchange in Multi-GPU programs with MPI.