ryan/mozjpeg
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

ChangeLog.md: clarification regarding AVX2

Browse Source
This commit is contained in:
DRC
2018-03-22 15:56:09 -05:00
parent 5b177b3cab
commit 2e90ccac0b
1 changed files with 6 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
ChangeLog.md
View File

@@ -6,11 +6,12 @@
1. Added AVX2 SIMD implementations of the colorspace conversion, chroma 1. Added AVX2 SIMD implementations of the colorspace conversion, chroma
downsampling and upsampling, integer quantization and sample conversion, and downsampling and upsampling, integer quantization and sample conversion, and
slow integer DCT/IDCT algorithms. When using the slow integer DCT/IDCT slow integer DCT/IDCT algorithms. When using the slow integer DCT/IDCT
algorithms, the compression of RGB images is approximately 13-36% (avg. 22%) algorithms on AVX2-equipped CPUs, the compression of RGB images is
faster (relative to libjpeg-turbo 1.5.x) with 64-bit code and 11-21% (avg. 17%) approximately 13-36% (avg. 22%) faster (relative to libjpeg-turbo 1.5.x) with
faster with 32-bit code, and the decompression of RGB images is approximately 64-bit code and 11-21% (avg. 17%) faster with 32-bit code, and the
9-35% (avg. 17%) faster with 64-bit code and 7-17% (avg. 12%) faster with decompression of RGB images is approximately 9-35% (avg. 17%) faster with
32-bit code. (As tested on a 3 GHz Intel Core i7. Actual mileage may vary.) 64-bit code and 7-17% (avg. 12%) faster with 32-bit code. (As tested on a
3 GHz Intel Core i7. Actual mileage may vary.)
2. Overhauled the build system to use CMake on all platforms, and removed the 2. Overhauled the build system to use CMake on all platforms, and removed the
autotools-based build system. This decision resulted from extensive autotools-based build system. This decision resulted from extensive