Sign up for the AI Accelerated Spark Challenge, June 23–30

Collaborate and build agentic applications with machine learning and large language models, or LLMs — powered by ASU’s Sol supercomputer and NVIDIA’s accelerated computing platforms.

Challenge statement

Hosted virtually on Zoom, the AI Accelerated Spark Challenge is a weeklong hackathon inviting teams to accelerate machine learning models using graphics processing units, or GPUs, harness the power of LLMs through interactive notebooks and design an innovative tutoring system using NVIDIA’s Python workflows and powered by ASU’s Sol supercomputer, comprising over 21,000 central processing units and over 290 NVIDIA A100 GPUs. Use datasets and tools provided by challenge organizers, or bring your own data science or LLM-related application idea to this challenge!

Teams start the challenge with a two-day bootcamp led by NVIDIA, ASU Research Computing and ASU Enterprise Technology, then innovate around the theme of personalized learning with artificial intelligence, or AI. Projects will be built and tested on ASU’s high-performance compute cluster, giving you real-world experience in scaling machine learning workflows. In addition to training support and access to Sol’s GPU partitions, teams will benefit from access to ASU’s MyAI Builder to support accelerated ideation, research and iteration throughout the challenge. Developed by Enterprise Technology’s AI Acceleration team, MyAI Builder is a tool offering ways to build custom AI experiences within ASU’s walled garden environment in as few as three steps, enabling innovation at scale. 

AI Accelerated Spark Challenge
Monday, June 23, 2025
9 a.m.–12:30 p.m.
Online via Zoom
Register to participate