Katrina Passarella-ward
Staff Software Engineer And Manager at Google
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolKatrina Passarella-ward is a Staff Software Engineer and manager at Google in San Francisco with six years of industry experience building high-performance imaging and computer vision systems for mobile devices. She advanced from hands-on engineer to architect and leader, shipping optimized algorithms that powered Pixel camera autofocus and contributed to industry-leading DxOMark scores. Katrina pairs low-level performance engineering—evidenced by contributions to the Halide compiler and Hexagon HVX runtime for correct and efficient bit-shift operations—with team leadership across cross-functional software-hardware projects. Her background in applied mathematics and computer science from UC Berkeley underpins a pragmatic, measurement-driven approach to optimization and system design. Colleagues rely on her ability to mentor engineers while driving complex, hardware-aware compiler and runtime changes that rarely show up on the surface but materially improve product performance.
6 years of coding experience
7 years of employment as a software developer
Bachelor of Arts (B.A.), Applied Mathematics and Computer Science, Bachelor of Arts (B.A.), Applied Mathematics and Computer Science at University of California, BerkeleyKamiak High School
English, Spanish
Github Skills (15)
performance-tuning10
compiler10
hexagons10
c-language10
performance-monitor10
compiler-compiler10
performance-analysis10
performance-monitoring10
performance-analytics10
cprogramming-language10
performance-measurement10
llvm9
gpu8
dsl7
image-processing5
Programming languages (1)
Github contributions (1)
halide/Halide
Aug 2019 - Jan 2020
a language for fast, portable data-parallel computation
Role in this project:
Backend Developer & Performance Engineer
Backend Developer & Performance EngineerContributions:9 commits, 4 PRs, 18 pushes in 5 months
Contributions summary:Katrina primarily contributed to improving the Halide compiler's performance and functionality, specifically focusing on optimizing bit shift operations within the Hexagon backend. They added support for negative bit shifts, ensuring the correct handling of these operations on the Hexagon architecture. Furthermore, the user modified and updated the HVX (Hexagon Vector eXtensions) runtime, including changes to assembly files (.ll) for enhanced efficiency and accurate operation of these bitwise functions. These changes reflect a focus on low-level optimization and hardware-specific code generation.
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.