3 ms·

> Rust is also the fastest in half of… Not true. That page only shows measurements for Rust and C (GCC) programs. Other programming languages are shown in th

by igouy 7y ago

> Rust is also the fastest in half of… Not true. That page only shows measurements for Rust and C (GCC) programs. Other programming languages are shown in the benchmarks game: 1) For binary-trees, there's a Pascal program that's faster — https://benchmarksgame-team.pages.debian.net/benchmarksgame/performance/binarytrees.html https://benchmarksgame-team.pages.debian.net/benchmarksgame/... 2) For reverse-complement, there's a C++ program that's faster — https://benchmarksgame-team.pages.debian.net/benchmarksgame/performance/revcomp.html https://benchmarksgame-team.pages.debian.net/benchmarksgame/... 3) For pidigits, there's a Chapel program that's faster — https://benchmarksgame-team.pages.debian.net/benchmarksgame/performance/pidigits.html https://benchmarksgame-team.pages.debian.net/benchmarksgame/... ---- Currently a Rust program is the fastest nbody program. Currently a Rust program is the fastest in 1 out of 10 …

altmind 7y ago
You are technically correct. Ok, replace "half" with "quarter, and the rest is within 30% of the fastest impl"
- igouy 7y ago
  No, not technically correct — plain vanilla indisputably correct. > Ok, replace "half" with "quarter So you claim — "Rust is also the fastest in quarter of the tests" ? Which 2.5 tests ? There are 4 other programs that the Rust spectral-norm program is not faster than. > "the rest is within 30% of the fastest impl" Not true. The Rust k-nucleotide program is not within 30% of the fastest program. The Rust regex-redux program is not within 30% of the fastest program.