MLCAD26-Contest
LLM-based Algorithm Discovery for Timing Optimization
Join the exciting MLCAD 2026 Contest and showcase your innovative skills in combining agentic Large Language Model (LLM) flows with Electronic Design Automation (EDA). This year’s challenge focuses on algorithm discovery using LLMs for timing optimization.
Winners of this year’s contest will be invited to MLCAD2026 to present their solutions as a paper in the proceedings.
Contest Overview
The goal of this contest is to create a design-aware timing optimization tool. Specifically, this means you can use an LLM to update/generate/modify the EDA tool’s source code and/or internal algorithms, depending on the design to which it is applied. Contestants may begin with the baseline OpenROAD source code.
Participants may use LLMs, agentic workflows, or other methods to develop new algorithms for timing optimization and legalization on a per-design basis
Contest Challenge
- Designing a design-adaptive timing optimization algorithm that dynamically modifies strategies based on the characteristics of each design.
- Applying LLM-based or agentic optimization workflows to discover, generate, or evolve timing optimization.
- Demonstrating improved timing closure and overall design quality after placement legalization and global routing, while balancing power, routability, and runtime constraints.
Why Participate?
- Showcase your innovative solutions at MLCAD 2026. Winners will have their papers invited to the MLCAD 2026 Proceedings!
- Apply for travel grants (available for eligible participants) to MLCAD 2026.
- Winners will receive high-performance NVIDIA GPUs as prizes!
Registration, Participants List, and Important Dates
Registration
- Registration opens : March 1, 2026
- Registration deadline: April 10, 2026
Registration Closed
Participants List:
If you have any questions about registration, please get in touch with us via the email address provided at the end of the page.
| Team ID | Team Name | Affiliation |
|---|---|---|
| team1 | Team Aliens | Arizona State University* |
| team2 | TDA | InfiniTree |
| team3 | one person one chip | Arizona State University* |
| team4 | Heuristic Horizons | N/A |
| team5 | ECO | University of California, San Diego |
| team6 | G.H.O.S.T | New Mexico State University |
| team7 | CODA | College of Integrated Circuits & Micro-Nano Electronics, Fudan University |
| team8 | DACCS | University of Southern California |
| team9 | AIDA | Kyung Hee University |
| team10 | LearningLLM | LTH |
| team11 | UBC_Fun_Team! | University of British Columbia |
| team12 | TimingCrafter | Shanghai Jiao Tong University |
| team13 | SSS | IMEC |
| team14 | TcT | National Tsing Hua University |
| team15 | justmekv | Arizona State University* |
| team16 | Makeit | Tongji University |
| team17 | SEUTiming | Southeast University |
| team18 | LAMDA-EDA | Nanjing University & Huawei Noah’s Ark Lab |
| team19 | Slug-CAD | UC Santa Cruz |
| team20 | APP | Peking University, Beijing, China |
| team21 | curetiming | CUHK |
| team22 | GGBond | Southeast University |
| team23 | masc-it | UC Santa Cruz |
| team24 | ItJustWorks | ShanghaiTech University |
| team25 | Lunaris | Jilin University |
| team26 | OpenPath | Wuhan University |
| team27 | have a try | Sun Yat-Sen University |
| team28 | 努力做竞赛队 | Southeast University |
| team29 | ACE-2 | NDSU |
| team30 | SeDe | Ulsan National Institute of Science and Technology |
| team31 | GoodSeDa | Ulsan National Institute of Science and Technology |
| team32 | Krusty Krab | Ulsan National Institute of Science and Technology |
| team33 | ALAP | Texas A&M University |
| team34 | Niuniu | The Hong Kong University of Science and Technology (Guangzhou) |
| team35 | AI4SEMI | SSIR |
| team36 | agentic-llm-for-eda | Texas A&M University |
| team37 | AVS | Indian Institute of Technology Guwahati |
| team38 | Luffy | JLU |
| team39 | xfunc | Functech |
| team40 | SMTraining | Sookmyung Women’s University |
| team41 | UT^2 | The University of Tokyo and University of Tsukuba |
| team42 | Thunder | Westlake University |
| team43 | Illegal Placement | International Institute of Information Technology Hyderabad |
| team44 | Mystic Musketeer | Khulna University of Engineering & Technology |
| team45 | Approx. Noumena | Barcelona Supercomputing Center |
| team46 | Bad Timing | Arizona State University* |
Note: Registrations closed !! Participants’ list fully updated with all registrations till 04/10/2026. Reach out to the provided email if you need to update your registration. You won’t receive a “registration successful” email when you submit your form, so please follow the website instructions to verify your participation.
Contest Details
-
Repository for the benchmark and evaluation scripts can be accessed here MLCAD26-Contest-Scripts-Benchmarks.
-
Please find the contest description: https://github.com/ASU-VDA-Lab/MLCAD26-Contest-Scripts-Benchmarks/blob/main/MLCAD2026-Contest-Algorithm-Discovery.pdf
Contest Timeline
| Phase | Milestone | Date |
|---|---|---|
| Announcement | Contest Announcement | March 1, 2026 |
| Benchmark Release | Initial benchmark suite | March 4, 2026 |
| Registration | Registration deadline | April 10, 2026 |
| Alpha Phase | Alpha submission | April 30, 2026 |
| Beta Phase | Beta submission – Testcase 1 release | May 5, 2026 |
| Beta Phase | Beta submission – Testcase 1 deadline and Testcase 2 release | May 15, 2026 |
| Beta Phase | Beta submission – Testcase 2 deadline and Testcase 3 release | May 25, 2026 |
| Beta Phase | Beta submission – Testcase 3 deadline | June 5, 2026 |
| Results | Results announcement | June 15, 2026 |
| Publication | Camera-ready paper | July 25, 2026 |
Prizes
Top-performing team(s) will be invited to present their solutions at MLCAD2026 and will be invited to submit a paper in the proceedings.
About MLCAD
The International Workshop on Machine Learning for CAD (MLCAD) is the leading venue dedicated to advancing research at the intersection of machine learning and electronic design automation (EDA). It provides a unique platform for collaboration between academia and industry, fostering innovation and driving progress in AI-driven CAD solutions.
Join the contest, push the boundaries of EDA, and lead the future of AI in chip design!
Acknowledgement
Thanks to:
- UCSD and the NSF POSE: Phase II: Growing a Collaborative Ecosystem for Open-Source Chip Design Using OpenROAD
- Purdue University and the NSF Chipshub
for sponsoring this contest and their involvement in organizing it.
Contest organizers
| Name | Affiliation |
|---|---|
| Atmadip Dey | ASU |
| Taizun Jafri | ASU |
| Janakiraman Ethirajulu | ASU |
| Vidya A Chhabria | ASU |
For questions, reach out to mlcad2026-contest@googlegroups.com
*Please note that these teams are not from the research group organizing the contest. Due to conflicts of interest, teams affiliated with ASU can participate but will not be eligible for prizes.