GenAI Chip Hackathon: Tackling Chip Design Challenges with Generative AI

The ICLAD-DAC 2026 GenAI Chip Hackathon is a competitive research and educational event co-located with the Design Automation Conference. Participants will use Generative AI models to solve chip design tasks across the RTL-to-GDSII flow, including RTL design, verification, and physical design.

The hackathon includes two tracks:

·  On-Prem Track: Participants use resource-constrained local models, including SLMs running on laptops.

·  Cloud Track: Participants may use cloud-hosted models, including large-scale LLMs.

Both tracks include an in-person competition at DAC 2026 in Long Beach on Sunday, July 26, 2026. Participants must be available to attend in person on this date to compete for prizes. Participants must also register for DAC, or I LOVE DAC to access the in-person event.

The 2026 hackathon has a hybrid structure. The On-Prem Track is primarily an in-person competition, with pre-DAC preparation resources provided to support teams. The Cloud Track includes both a pre-DAC online phase and an in-person phase at DAC.

All registered teams may attend DAC in person and compete on the day of the hackathon, subject to DAC or I LOVE DAC registration requirements.

For the Cloud Track, up to 10 presentation slots will be available at DAC. These presentation slots will be offered to teams based on their “Before DAC” online performance. Teams not selected for presentation slots may still participate in-person “On the Day of DAC” and remain eligible for prizes.

The hackathon will include multiple problem categories across the RTL-to-GDSII flow. Problems will vary in difficulty and may include tasks in RTL generation and optimization and layout.

Evaluation criteria will depend on the problem category and may include:

·  Correctness of the generated solution

·  Quality of design metrics such as power, performance, area, and timing

·  Ability to handle hidden test cases and benchmark designs

·  Automation quality and robustness of the proposed approach

·  Token cost, runtime, or resource usage where applicable

For both tracks, hidden benchmark problems or hidden test cases will be introduced in-person on the day of DAC.

The following infrastructure will be provided to support participation in the two tracks:

·  On-Prem Track: Qualcomm AI laptops with pre-loaded local models and EDA tools will be provided in-person on the day of DAC. Preparation access to Qualcomm Developer Cloud (QDC) will be provided before DAC.

·  Cloud Track: Google Cloud Platform access, GenAI model credits, and setup instructions will be provided prior to the contest stand for both the Before DAC and the On The Day Of DAC  phases of the competition.

Separate prizes will be awarded for the On-Prem Track and the Cloud Track.

·  On-Prem Track: Top-performing teams will receive an award. The winning On-Prem Track team will also get to keep the Qualcomm AI laptops used in the competition.

·  Cloud Track: Top-performing teams will receive an award.

Cloud Track winners will be selected based on combined performance from the Before DAC online phase and the in-person On The Day Of DAC phase. On-Prem Track winners will be selected based on the in-person DAC competition.

Participants may participate in the On-Prem Track, the Cloud Track, or both tracks.

Participation in the in-person DAC competition on Sunday, July 26, 2026, is required for prize eligibility. Participants must be physically present at DAC 2026 in Long Beach and must have DAC or I LOVE DAC access for the event.

Note that In-person participation at DAC is mandatory to be eligible for the awards.

Registration deadline: Jun 8, 2026

Phase 1: Remote Competition: June 8-July 10

Phase 2: In-Person @ Design Automation Conference: July 26, 2026

https://form.jotform.com/261064280546152