Summary of Changes
This update introduces enhanced documentation clarifying advanced build-time optimisation strategies for Revolution on modern x86-64 platforms. Specifically, the README has been extended to provide explicit guidance on:
- Enabling SSE4.1 and POPCNT prefetch support for
x86-64-bmi2builds, with an emphasis on architectures where these instruction sets are available and beneficial. - Layering Link Time Optimisation (LTO) on top of existing compilation flags to maximise instruction-level and cross-translation-unit optimisations.
- Establishing a methodological benchmarking approach, encouraging comparative performance evaluation against a clearly defined baseline build in order to quantify the practical gains of these optimisations.
The intent of these additions is to improve reproducibility, transparency, and performance literacy for developers and testers who compile Revolution with hardware-specific tuning, particularly in competitive computer-chess testing environments.
Testing
⚠️ Not executed.
This change is documentation-only and does not modify the engine’s source code or runtime behaviour. No functional or regression testing was required.