Optimize Rust backtesting hot paths with zero-allocation and SIMD techniques.
npx skills add https://github.com/gtnix/quant_b3_backtest --skill quant-engineer
Please help me install this Agent Skill. Skill: quant-engineer Source: https://github.com/gtnix/quant_b3_backtest/tree/main/.cursor/skills/quant-engineer Command: npx skills add https://github.com/gtnix/quant_b3_backtest --skill quant-engineer
Rust performance engineering for low-latency backtesting systems to squeeze out maximum efficiency and determinism.
Run a baseline benchmark and begin profiling hot paths to identify optimization targets.
High-intent search queries and answers about installing and using this skill.
Achieve deterministic Rust backtesting by applying fixed-point arithmetic instead of floating-point operations to ensure consistent numerical results. This method maintains deterministic simulation across different CPU architectures without introducing rounding discrepancies.
Profile Rust backtest performance regressions by establishing baseline benchmarks and profiling hot paths to identify optimization targets. This discipline creates performance regression guardrails to ensure rigorous profiling standards are maintained.
SIMD vectorization is supported for low-latency backtesting workloads in Rust through architecture-aware memory layout optimization. This technique maximizes CPU throughput by processing multiple data points simultaneously within the backtesting engine.
Non-deterministic Rust backtest results across architectures often stem from floating-point rounding inconsistencies. Applying fixed-point arithmetic enforces deterministic simulation, ensuring identical numerical outputs regardless of the underlying hardware architecture.
JylhisProfile Go hot paths with pprof and validate optimizations using benchstat.
yashurathodOptimize v3 performance bottlenecks in Flash Attention, HNSW indexing, and memory usage.
yfyang86Analyze numerical computation problems using IEEE-754 floating-point principles.
jmdueaGuide embedded Rust design with no_std, heapless, and interrupt-safe patterns.
