The IDRIS Open Hackathon is a multi-day, intensive event designed to assist computational scientists and researchers in porting, accelerating, and optimizing their applications for various data center architectures, including CPUs and GPUs. This event provides participants with access to Dalia, Europe’s first NVIDIA GB200 NVL72 supercomputer for open research, which integrates 36 NVIDIA Grace CPUs and 72 NVIDIA Blackwell GPUs. Throughout the hackathon, teams are paired with dedicated, experienced mentors in programming and specific application areas to help them achieve significant performance gains and speedups using a range of programming models, libraries, and tools.
Event Schedule and Key Dates
The IDRIS Open Hackathon schedule is organized into a pre-event onboarding session followed by the main event period. It includes an online team and mentor introduction meeting on May 5, then online and hybrid technical working sessions during the listed event dates, and concludes with final presentations. Key dates and daily milestones are listed below.
• Application Deadline: March 10, 2026
• Event Dates: May 12-21, 2026
• Day 0 (Online): May 5, 2026, 10:00 AM - 11:30 AM CEST - Teams and Mentors introduction meeting.
• Day 1 (Online): May 12, 2026, 09:00 AM - 05:00 PM CEST - Includes welcome, cluster introductions, team presentations, and breakout sessions.
• Day 2 (Hybrid): May 19, 2026, 09:00 AM - 05:00 PM CEST - Features daily scrums and dedicated time for teams to work on codes with mentors.
• Day 3 (Hybrid): May 20, 2026, 09:00 AM - 05:00 PM CEST - Continues with daily scrums and mentor-supported code development.
• Day 4 (Hybrid): May 21, 2026, 09:00 AM - 05:00 PM CEST - Concludes with final code work and final team presentations.
Logistics and Format
• Format: The IDRIS Open Hackathon is a hybrid event. Days 0 and 1 (May 5 and May 12) are fully online. Days 2-4 (May 19-21) are held in-person at IDRIS-CNRS or online. The official time zone is CEST.
• Venue: IDRIS-CNRS (campus of the Paris-Saclay University), Rue John Von Neumann, Bâtiment 506, BP 167, 91403 ORSAY CEDEX.
• Attendance: In-person attendance is preferred. It is mandatory for at least two members of each team to be onsite at IDRIS.
• Expenses: Attendees are responsible for covering their own travel and accommodation expenses. Logistical information and hotel suggestions will be provided to accepted teams.
• Communication: All official communication for both remote and in-person participants will be conducted via Zoom, Slack, and email.
Participation Requirements
• Team Prerequisites: Teams are expected to be fluent with the code or project they bring and be motivated to make significant progress during the event.
• Attendance Policy: Full participation is required. A lack of complete participation may result in the entire team being excused from the event.
• Project Licensing: Projects must have a license attached and detailed in the application. Permissive-style Open-Source Licenses (e.g., BSD, MIT, Apache 2.0) are preferred to facilitate mentor pairing.
Preparation and Resources
• Compute Resources: All attendees will be granted access to a GPU cluster for the duration of the hackathon.
• Recommended Skills: While no advanced parallel computing or GPU skills are required, it is helpful for teams to have a basic understanding of GPU programming and profiling. Organizers can provide access to no-charge lectures, tutorials, and labs to help participants prepare.
• Post-Acceptance: Accepted teams will receive registration information, mentor introductions, and cluster access details. All team members must register for the event and review the provided Attendee Guide.
The Open Hackathon program, launched in 2014, connects scientists, researchers, and developers with the latest technologies and collective expertise to accelerate and optimize their projects. By partnering with foremost research institutions and high-performance computing centers, the program provides the research and developer community with a unique opportunity to build the skills needed to leverage best-in-class computing systems, enabling them to advance critical scientific discoveries.