Princeton Ferro

Email: princetonferro@gmail.com
GitHub: https://github.com/Prince781
LinkedIn: https://www.linkedin.com/in/princeton-ferro/

Interests

compilers, languages, runtimes, toolchains, operating systems, parallel programming

Skills

Languages

• C, Python, Vala (great)
• C++, C#, JavaScript, TypeScript (good)
• Java (familiar with, but haven’t touched in awhile)

Tools

vim, gdb, rr, valgrind, perf, git, meson, make

Institutions

New York University (2019-2022)
Masters in Computer Science

University of Rochester (2015-2019)
Bachelor of Science, Computer Science

Bergen County Academies (2011-2015)
High School diploma

Projects

Vala

• Vala is a high-level programming language that compiles to C which then compiles to native code. It takes inspiration from C# and other languages. Vala has been used to write many applications for Linux and is used extensively by the elementaryOS and GNOME projects/desktop environments.
• I’m the main developer of the Vala Language Server, as well as a templating system for Vala projects, and I hack on the Vala compiler. I also contribute a bit to the rest of the GNOME stack.
• Vala Language Server has over 200 stars, and the Visual Studio plugin has over 9k installs. The language server is officially supported in GNOME Builder, the official IDE of GNOME

LSTF

• A domain-specific programming language for testing language servers. Still a work in progress.
• The language is based on JSON data types, with closures and asynchronous programming thrown in the mix.
• The implementation has an optimizing compiler, stack-based virtual machine, and language runtime.
• Source: https://github.com/Prince781/lstf

Cache-Conscious Hash Array-Mapped Trie for Rust (2018)

• Alternative to std::collections::HashMap that uses a lock-free trie with a contiguous allocator for elements, improving scalability with multiple threads.
• Source: https://github.com/chichunchen/concurrent-cache-conscious-hamt-in-rust

Dynamic GPU Acceleration of Linear Algebra Computations (2017-2019)

• A drop-in replacement for BLAS and LAPACK that interops with cuBLAS or clBLAS to implement GPU acceleration in scientific applications.
• Paper: http://prince781.net/files/dyn_gpu_accel_lin_alg.pdf

Research

Managing Application Parallelism through Parallel Efficiency Regulation (2018-2021)

• A daemon that optimizes system resource allocation by monitoring per-application performance metrics using hardware counters.
• Poster: https://prince781.net/files/mapper-ppopp-poster2019.pdf
• Paper: https://prince781.net/files/mapper-taco2022.pdf
• Source: https://github.com/URCS-systems/MAPPER/tree/release

Employment

NVIDIA (June 2022 - present)

Compiler Development Engineer - Remote / NJ / Santa Clara, CA

• Work on compiler optimizations for GPU compute (CUDA, OpenCL)

L3Harris Technologies (August - September 2020)

Intern - Space and Airborne Systems, Clifton, NJ

• Implemented a test framework in Python for flight and mission control software

Cray (now HPE) (May - August 2019)

Compiler Software Intern, Bloomington, MN

• Improved OpenMP performance in Cray’s clang+LLVM fork with optimizations for OpenMP fork/join
  
• Improved debugging support with Cray’s OpenMP runtime libcraymp
  
• Poster: http://prince781.net/files/cray-poster.pdf

BrightLogic (October 2014 - August 2015, June - August 2017)

Software Engineer Intern, Midland Park, NJ

• Developed a GUI for monitoring hardware state of the ActiveFloor (a floor with interactive panels)
  
• Developed a video game for the ActiveFloor
• Created a centralized configuration manager for various software components, using Angular and C#
  
• Created libraries and tools for ActiveFloor software
  
• Rapidly prototyped demo apps for the ActiveFloor to show to customers

Teaching

Teacher’s Assistant - Programming Language Design and Implementation (September 2018 - December 2018)

University of Rochester

• Drafted and graded assignments, tests, and quizzes. Held office hours with students.