This is a custom Linux kernel built specifically for AI workloads as part of The Capstone Project. The goal was to strip out unnecessary features, reduce system overhead, and optimize performance for tasks like machine learning and data processing on modern hardware.
- Faster Startup: Kernel compression switched to LZ4, trading a bit of compression for quicker boot times.
- Simplified Features: Removed legacy functionality (like old hardware emulation and outdated system calls) to make the system leaner and more efficient.
- Optimized for Modern PCs: Stripped out support for hardware that's currently never used, like non-PC x86 platforms and configurations meant for servers or older systems.
- Better Memory Handling: Tweaked memory settings to better support AI workloads without bloating the system.
For more details on changes read the .changes file.
For more details on the kernel, visit https://docs.kernel.org