I'm Anthony, a computer science student at Carnegie Mellon University. I'm interested in C++ and parallel computing for performance optimization, compilers and machine learning. Feel free to reach out to me at [email protected]!

In the meantime, check out some of my projects:

• Mirage Persistent Kernel⚡: The research project I'm currently working on. It's a compiler from a LLM model to a single GPU kernel that uses fine-grained data dependency, scheduling, and computation overlap to optimize for GPU inference latency. I'm working on shared memory reuse and software pipelining between atomic subtasks.
• CSerpent🐍: A compiler for a typed Pythonic language to native C++, aiming for an order-of-magnitude speedup. I wrote both the performant runtime system in C++ as well as the backend semantic routines for code generation. We're still looking for help, so reach out!
• CMR Driverless Controls🚗: As captain of the controls team, we've developed a state-of-the-art model predictive controller (MPC) for our Formula SAE driverless racecar, 22a. It uses the MPPI algorithm's tremendous parallelizability to simulate a million trajectories per second to gun the car at 30mph on an unseen track. This year, we're aiming to make it even faster. I also prepared some documentation in hopes of benefitting other driverless teams.
• nnUNet👁️: During an internship with A*STAR, I added semi-supervised training onto the nnUNet framework to make use of unlabeled medical imaging data. This work was based off version 1 of nnUNet, hence the fork from that branch.
• Automajumper🎮: A Unity game I made with my friends from CMU's Game Creation Society. It's a rhythm platformer based on Conway's Game of Life. I wanted to do something other than code, so I took charge of Level Design instead. You can play the game here!
• Booth👋: As Electricals chair for the Singapore Student's Association's Spring Carnival Booth, we made a fully-functional claw machine spanning the 2.4m wide booth using magnets an an Arduino.