Nemotron Labs

A live look at how Motif, a small Korean team, achieved a top 5 AA II score using NVIDIA’s technology stack.

Discover how Motif, a small Korean AI team, achieved a top-five AA II score using the NVIDIA technology stack. This NVIDIA Nemotron Labs session explores the technical strategies behind Motif’s performance gains, the role NVIDIA technologies played, and the trade-offs the team navigated with limited resources. Attendees will gain practical insights for building world-class AI systems without big-tech scale, and see how Motif’s success contributes to Korea’s broader sovereign AI ambitions.

What you'll learn:

• The technical approach Motif took to push their AA score higher
• How NVIDIA’s stack contributed to that performance gain
• Key trade-offs a resource-constrained team had to n

Running AI on local devices is becoming more valuable over time. As models get more capable and hardware gets more powerful, local AI is unlocking new possibilities: sovereignty, privacy, cost efficiency, resilience, and access in regulated or air-gapped environments.
In this session, Exo Labs joins Nemotron Labs for a live demo of Nemotron 3.5 Lightning on DGX Spark and a discussion on why local AI and why now.
What you'll learn:

• Why local AI has crossed from interesting to useful and why its value keeps compounding
• What open models like Nemotron 3.5 Lightning unlock for developers running AI on their own hardware
• How DGX Spark and DGX Station fit into a local AI stack built for always-on agentic workloads

Have questions about local AI or running Nemotron on DGX Spark or DGX Station? Drop them in chat live, and our guests will answer in real time.