HackerTube 中文?

CppCon 2015: Chandler Carruth "Tuning C++: Benchmarks, and CPUs, and Compilers! Oh My!"

CppConYouTube ↗

Further reading 1

When one component fails, who can keep working? →One process fails. Must unrelated requests wait? Separate failure propagation from throughput limits, then trace what the operating system and service manager can actually isolate.This video helps answer · Did the timed loop actually do the work?

Caption notes

Carruth follows a vector microbenchmark down to the generated instructions. If the compiler cannot observe the intended work, the timed loop may measure something quite different from the source-level operation.

Video or transcript source ↗
Original passages · 2
micro benchmarks are when you're benchmarking some really isolated part of your system
CppCon 2015: Chandler Carruth "Tuning C++: Benchmarks, and CPUs, and Compilers! Oh My!" · ▶ 10:51–10:56
In my case the observable side effect is the benchmarking.
CppCon 2015: Chandler Carruth "Tuning C++: Benchmarks, and CPUs, and Compilers! Oh My!" · ▶ 44:30–44:34

Hacker News comments

View original on HN ↗