Zvi Rackover is a systems software manager and engineer with over a decade of experience building high-performance compilers and low-level software for ARM64 and x86/AVX2 architectures. He leads teams at Apple while maintaining deep hands-on expertise from his long tenure at Intel working on OpenCL SDKs and compiler engineering. Zvi’s contributions to prominent open-source projects like Neovim and Halide demonstrate a mix of practical developer tooling improvements and performance-focused compiler optimizations. He combines multi-programming and SoC knowledge to bridge hardware-aware code generation with production-grade runtimes. Based in Israel and holding a BSc in Electrical Engineering from Technion, he’s as comfortable driving architecture and management decisions as he is adding targeted tests and removing legacy cruft in complex codebases. An understated strength is his ability to surface measurable performance wins through careful low-level pattern and intrinsic work that scales across platforms.
11 years of coding experience
5 years of employment as a software developer
BSc., Electrical Engineering, BSc., Electrical Engineering at Technion - Israel Institute of Technology
Contributions:18 commits, 7 PRs, 42 comments in 2 months
Contributions summary:Zvi contributed to the Neovim project by primarily improving the statusline functionality and removing deprecated code. They added tests for statusline expressions and made changes to the buffer and regexp functionalities within the project. Additionally, they refactored the code by removing legacy and deprecated functions.
a language for fast, portable data-parallel computation
Role in this project:
Performance Engineer
Contributions:6 commits, 6 PRs, 18 comments in 1 day
Contributions summary:Zvi focused on optimizing the Halide compiler for x86 architectures, specifically targeting AVX2 instruction sets. They implemented and tested new runtime functions and patterns to leverage AVX2 for image processing tasks. Their contributions involved adding missing code generation patterns and strengthening existing tests to ensure the correct usage of AVX2 intrinsics within the Halide compiler, leading to improved performance.
computationhexagonhalideparallelgpu
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.