Over a two-month period, contributed to NVIDIA/CUDALibrarySamples by developing and integrating distributed FFT planning samples for cuFFT MP, enabling multi-GPU and multi-rank FFT workloads within a unified repository. The work involved implementing demonstration samples that showcased both automatic and custom data distribution strategies using C++ and CUDA, with MPI facilitating distributed computation. Additionally, stabilized the codebase by reverting unintended changes to sample directories and correcting the MPI communicator attachment in reshape operations. These efforts improved the repository’s support for high-performance computing workflows, allowing users to evaluate and integrate distributed FFT planning with cuFFT MP in real-world scenarios.