CachyOS: x86-64-v3 Microarchitecture Optimization
CachyOS is committed to delivering high-performance computing to its users. Our installer detects the microarchitecture of your machine and uses optimized packages for x86-64-v3 systems, resulting in a significant performance improvement of more than 10%.
Optimized Packages
All packages in CachyOS are compiled with Link Time Optimization (LTO) and optimized for the x86-64-v3 microarchitecture. We are constantly working on expanding our repositories and adding a x86-64-v4 repository which will be filled over time. The GCC compiler is optimized using Profile Guided Optimization (PGO) and LTO, and certain packages are further optimized with BOLT, a post-link optimizer.
Performance Enhancements
CachyOS comes with several performance enhancements. These enhancements are designed to improve the overall user experience. These include
- Ananicy-CPP, which is preinstalled and configured by default.
- ZRAM, which is pre-configured with zstd compression.
- The BFQ scheduler, which is used by default for HDDs and mq-deadline for SSDs.
- Highly optimized, patched and tuned Kernel.
- BBRv2 set as the default network setting.
- Additional tweaks and improvements to further optimize performance.
Security Features
At CachyOS, we take security very seriously. That is why our software includes the following security features:
- A Firefox-based browser with security plugins and additional features
- The option to use Firejail, which runs untrusted applications in a sandbox environment and updates automatically with new packages
- Preconfigured DNS servers that utilize DoH and DoT protocols
- Anonymized DNSCrypt configuration and use of only DNSCrypt servers
- A hardened kernel available in the CachyOS repository
- Security-related packages compiled with extra CC flags for added security
Ananicy-Cpp
Ananicy-Cpp is a shell daemon that manages process IO and CPU priorities with community-driven rulesets for popular applications. This tool is primarily designed for desktop usage and is a complete rewrite of the original Ananicy in C++, providing lower CPU and memory usage. Additionally, Latency_nice has recently been included to help the scheduler make decisions about which tasks receive lower or higher latency.