This repository is a collection of experiments exploring kernel-level optimizations and scheduling strategies for High-Performance Computing (HPC) workloads.
These experiments aim to systematically evaluate how different Linux kernel configurations, scheduling policies, and system-level tuning affect performance across a variety of HPC workload types.
Current experiments include:
- Tickless Kernel Experiment
- Scheduler Policy Experiment on CFS
- Pre-emption model choice
More experiments will be added as this project evolves.