Core Topics
- Probabilistic and thermodynamic computing
- Energy-Based Models
- THRML framework
1-Day Hackathon
Build with probabilistic and thermodynamic computing.
A hands-on event for learning Energy-Based Models, the THRML framework, and emerging computing paradigms designed for the next generation of AI systems.
About the event
Educational deep-dive plus real hackathon.
The event introduces probabilistic computing chips, Energy-Based Models, and THRML through theory, guided practice, and a team-based hackathon.
Audience
- ETH students and researchers
- Industry professionals
- Teams of 1 to 4 participants
Optional preparation
Custom learning plan
Prepare before the event with a forkable plan you can adapt to your level, needs, and interests.
Sponsored by openLesson.academyAgenda
Starts at 12:30.
All times are CEST, Zurich time. Exact room location remains to be confirmed.
Lecture
Intro to probabilistic computing. Theory and Practice.
Extropic CEO Guillaume Verdon
Live or pre-recorded session with Extropic's CEO.
Hackathon
Teams of 1 to 4 build projects with organizer and mentor support.
Dinner break
Food, reset, and final project preparation.
Demos
Short demos from each team followed by judging.
Announce winners
Prize announcements, closing remarks, and next steps for the community.
Technology context
About Extropic
Extropic is building energy-efficient computers for the AI-powered future by rethinking computing from physics fundamentals.
Core focus
Thermodynamic and probabilistic computing for generative AI and probabilistic workloads, targeting orders-of-magnitude lower power for inference and generation.
Key technology
Thermodynamic Sampling Units, the XTR-0 / X0 platform, and THRML, an open-source Python library for thermodynamic algorithms and TSU simulation.
Education context
About openLesson
openLesson.academy is part of the uncertain.systems effort to accelerate technical education around emerging computing paradigms.
Education accelerationism
The goal is to shorten the path from frontier research to practical understanding: clear lessons, hands-on projects, and public learning infrastructure for students, researchers, and builders.
Why it matters
Probabilistic and thermodynamic computing will need a new generation of practitioners. openLesson supports this by turning complex ideas into learnable, buildable material.
Optional preparation
Probabilistic Computing and Extropic AI approach
Fork the custom learning plan and adapt it to your level, needs, and interests.
Open learning planSponsors and partners
Supported by
$3,000 prize pool
1st Prize: $1,500
2nd Prize: $1,000
3rd Prize: $500
Registration
Apply to join the hackathon.
Seats are limited to keep the event intimate and high-impact.
Important note
You have to bring your own laptop and coding environment. We will provide access to everything else, including libraries, simulators, and supporting materials.
Organizer
Daniel Colomer
uncertainsystems@gmail.com