Jaemin Park

Computer Science and Engineering · Seoul National University

profile.jpg

I'm a computer science and engineering student at Seoul National University. My interests span algorithms and data structures, computer systems, and machine learning.

My work ranges from geometric data structures and compiler optimization to hardware simulation. I build tools that make the behavior of processors, memory, and I/O devices observable, and help organize competitive programming practice.

Selected Achievements

  • 2024

    The 48th Annual World Finals of the International Collegiate Programming Contest

    9th place (Bronze Medal) · Team NewTrend

  • 2024

    The 2024 ICPC Asia Pacific Championship

    1st place (Champion) · Team NewTrend

  • 2023

    Samsung Collegiate Programming Cup 2023

    2nd Prize

Codeforces: International Grandmaster; rating 2715.

All awards and results

Experience

Huawei Technologies Co., Ltd. — Hong Kong Research Center

2025

Research Intern

Worked on an LLVM-based C++ framework for tensor computation graphs, contributing a compiler optimization pass and a DAG analysis tool based on flow algorithms.

Korean IOI Training Camp

2021 – 2024

Teaching Assistant

Assisted in teaching algorithms to more than 50 candidates for the Korean IOI delegation.

Projects

Reinventing the Wheel

A hardware simulation platform for computer architecture research, making processor, memory, and I/O behavior observable from gates to program execution.

The platform brings together ALUs, flip-flop/SRAM/DRAM memory models, caches, and RV32I/RV32IM processors. An event-driven C++20 engine combines structural and behavioral models with gate-derived delay tables and graph-based computation of safe clock periods.

Research

My earlier research explored geometric data structures for maintaining function order statistics and envelopes, dynamic programming optimization, and reinforcement learning for graph coloring games.

Research summaries

Blogs

I write contest solutions, mathematical proofs, and paper reviews in three collections: