From 7829d1a9f6d680aa93474ff45ad23de98f2aad35 Mon Sep 17 00:00:00 2001 From: Daniel Lemire Date: Mon, 19 Oct 2020 21:10:20 -0400 Subject: [PATCH] Update README.md --- README.md | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/README.md b/README.md index c0557e5..ec90f30 100644 --- a/README.md +++ b/README.md @@ -59,6 +59,19 @@ the type `fast_float::chars_format`. It is a bitset value: we check whether to determine whether we allow the fixed point and scientific notation respectively. The default is `fast_float::chars_format::general` which allows both `fixed` and `scientific`. +## How fast is it? + +We find that it is often twice as fast as the best available competitor, and many times faster than many standard-library implementations. + +``` +# parsing random integers in the range [0,1) +volume = 2.09808 MB +netlib : 294.33 MB/s (+/- 2.4 %) 14.03 Mint/s +strtod : 86.53 MB/s (+/- 1.4 %) 4.12 Mint/s +abseil : 521.66 MB/s (+/- 3.2 %) 24.86 Mint/s +fastfloat : 1061.86 MB/s (+/- 3.8 %) 50.61 Mint/s +``` + ## Using as a CMake dependency This library is header-only by design. The CMake file provides the `fast_float` target