Jonathan Wright
Principal Software Engineer at Arm
Greater Cambridge Area United Kingdom
Summary
👤
Senior🎓
Top SchoolJonathan Wright is a Principal Software Engineer at Arm with eight years of deep expertise in SIMD and Arm vector-extension optimizations, specializing in Neon, SVE and SVE2 for high-performance codecs and image processing. He has progressed through technical ranks at Arm from graduate to principal engineer, routinely delivering hand-tuned assembly and intrinsics rewrites that squeeze significant speedups on AArch64 using recent ARMv8.4/8.6 instructions like SDOT/USDOT. An active open-source maintainer and contributor, his performance work appears in widely used projects such as libvpx, libjpeg-turbo and Chromium, where his refactors often replace older GAS/C code with leaner Neon implementations. Jonathan combines systems-level discipline—evidenced by MISRA compliance and platform errata fixes in Trusted Firmware—with a practical propensity for manual unrolling and compiler-aware refactoring that yields measurable gains. Based in the Greater Cambridge area, he brings rare low-level optimization craft applied to mainstream multimedia stacks.
8 years of coding experience
7 years of employment as a software developer
AS/A Level, AS/A Level at Caerleon Comprehensive SchoolMA (Cantab) Computer Science, MA (Cantab) Computer Science at University of Cambridge
Github Skills (29)
assembly10
c-language10
image-processing10
image-optimization10
architecture10
computer-architecture10
cpu-architecture10
c1110
gcc10
c1710
arch10
sys10
performance-optimization10
assemble10
assembler10
Programming languages (3)
Github contributions (5)
libjpeg-turbo/libjpeg-turbo
Jul 2018 - Feb 2022
Main libjpeg-turbo repository
Role in this project:
Backend Developer & Optimization Engineer
Backend Developer & Optimization EngineerContributions:2 reviews, 11 PRs, 59 comments in 3 years 7 months
Contributions summary:Jonathan contributed to performance optimizations within the libjpeg-turbo library by implementing and refactoring Neon intrinsics. This involved rewriting existing functions for various tasks like colorspace conversion (RGB to YCbCr, YCbCr to RGB, and RGB to Grayscale), DCT, downsampling, and upsampling. The user's work focused on leveraging Arm's Neon instruction set to improve the speed of image processing operations by replacing the previous GAS (GNU Assembler) implementations and C scalar code with more performant Neon intrinsics.
libjpeg-turboturbolibjpeg
mozilla/mozjpeg
Jun 2018 - Sep 2021
Improved JPEG encoder.
Role in this project:Back-end Developer
Back-end DeveloperContributions:22 commits in 3 years 3 months
Contributions summary:Jonathan primarily contributes to the optimization and implementation of Neon intrinsics for the mozjpeg project, which is an improved JPEG encoder. Their work focuses on enhancing the performance of critical image processing functions, including RGB to YCbCr and YCbCr to RGB color space conversions, as well as fast and accurate inverse DCT computations. The user's commits demonstrate expertise in low-level optimization and SIMD programming techniques.
image-optimizationimprovedcodecgifencoder
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.