Implementation and Acceleration of Parallel Computing Algorithms in HPC Environment
Personal Project
Overview
Challenged sparse matrix computation acceleration assuming large-scale scientific computing in HPC (High-Performance Computing) environments.
Details
Implemented parallel computing algorithms in C/C++ utilizing both shared memory parallel processing (OpenMP) and distributed memory parallel processing (MPI).
As the first stage of implementation, mastered the fundamentals of parallel programming through parallelization of pi calculation using the Monte Carlo method.
Subsequently applied to sparse matrix computation, a typical challenge in computer science, directly tackling low-layer challenges such as data structure partitioning, minimization of inter-process communication overhead, and load balancing to achieve computational acceleration.