Compiler upgrades can change generated code, but their effect on a FIX parser depends on the exact source, flags, target CPU and workload. We have not published a reproducible run supporting the previously stated GCC 16 speed gains, so those figures and feature claims have been withdrawn.
Measurement plan
Pin the source commit, compiler builds, optimization and LTO flags, CPU model, operating system and FIX message corpus. Build the same parser with each compiler, then record both clean build time and repeated runtime latency and throughput. Publish the commands, raw samples and statistical summary before making a percentage comparison.
Interpretation
Keep build-time and runtime results separate. Report warm-up, CPU affinity and variance, and avoid attributing a difference to a compiler feature unless generated code or compiler diagnostics identify it.
