Winning submission for StartHack 2024: HPC optimized multi-GPU/CPU inference
-
Updated
Aug 20, 2024 - C
Winning submission for StartHack 2024: HPC optimized multi-GPU/CPU inference
Code, configs, and playbooks for The Login Node's HPC From Scratch series. Real hardware. No shortcuts.
High Performance Computing. Sparse matrix operations using multi-threading with OpenMP in C
This project is a multi-threads Agent-based simulation framework.
MPI-based parallel histogram program using broadcast, scatter, and reduction to compute global bin counts across multiple processes.
To associate your repository with the high-perofrmance-computing topic, visit your repo's landing page and select "manage topics."