![]() The lib is also faster than Intel Performance Primitives, low-level building blocks for image processing optimized for a wide range of Intel architectures. ![]() On average, Pillow-SIMD is currently resizing images 15 times faster than ImageMagick. I forked Pillow, a Python image processing library used by Uploadcare, and called the fork Pillow-SIMD. Right when I started off, we decided to open the source code. With optimizations, Uploadcare now needs six times fewer servers to handle its load than before. I wasn’t building it for fun: I work for Uploadcare and resizing images has always been a practical issue with on-the-fly image processing. I call it “the fastest resize” because every other library I could find and test was slower. Part 0 is intended to give you the big picture. It doesn’t seem a single article has enough room for that, so I’ll be releasing more. However, I love tech details, so I’m planning to talk over plenty of those. My warmest thanks to fellow Pillow contributors for their comments and motivation: Alex Clark, Hugo van Kemenade, Andrew Murray.Įven though this article is mostly for developers, I seek to keep it simple. I’d like to share my experience hoping I might motivate and inspire you to go optimize things. ![]() I made the fastest image resize for modern x86 processors. First published on June 20, 2017, in Insights Everything Insights Engineering News Product Culture Compliance The fastest production-ready image resize. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |