Google Benchmark
Google's C++ microbenchmark library provides automatic timing, statistics, parameterization, complexity analysis, counters, JSON output, and portable builds for tests and CI.
Why use it
Google Benchmark lets engine teams test assumptions about containers, math, serialization, allocation, and algorithms during technical prototyping, then detect small persistent regressions as development continues.
Where it fits
Use prototypes and playtests to find core experience problems while they are still cheap to fix.
What to check
Microbenchmarks are sensitive to CPU frequency, background load, compiler optimization, and cache state; repeat them on controlled machines and correlate with real-frame profiling before architectural decisions.