CTO and Co-founder at Ative o Fit | Computer Science Student at UFLA
Much like the nature of a fractal, my approach to software engineering is to build scalable, robust, and strictly deterministic systems grounded in precise fundamentals.
I maintain a public archive to document my research, architectural designs, and technical articles. It is a space where I mostly share knowledge.
Floatless: Q16.16 & Q32.32 Deterministic Math in C++ (128-bit safe) A C++ library for fixed-point arithmetic (Q16.16 and Q32.32), architected to guarantee strict determinism in cross-platform simulations by entirely suppressing the inconsistencies inherent to floating-point numbers.
- Strict Type Safety: Explicit constructors that block accidental float/double conversions, ensuring data integrity within the simulation.
- Safe Q32.32 Arithmetic: Native 128-bit support in GCC/Clang, implementing a manual 64x64->128-bit fallback to bypass overflow risks in 32-bit compilers (MSVC).
- High-Performance Trigonometry: Sine and Cosine functions mapped to 1024-entry look-up tables (LUT), refined via linear interpolation (lerp).
- Bitwise Integer Square Root: Highly optimized bit-shift algorithm, replacing the inefficiency of the Newton-Raphson method in a fixed-point context.
- Email: filipegarciar@gmail.com
- Instagram: @filipe.g_r