Scaling AI is more than just stacking GPUs, it's about making every dollar and watt count.

Chris Sosa, Director of Engineering at AMD, reveals how they’re pushing the limits of AI hardware efficiency.

Chris shares:

▫️ The biggest bottleneck in scaling AI workloads today

▫️ Why idle GPUs drain your budget

▫️ The struggle between training vs. inference at scale

▫️ How AMD’s ROCm stack is changing AI accessibility

▫️ His vision for smarter AI orchestration and platforms

A must-watch for engineers, AI builders, and tech leaders hungry for insights on the future of AI infrastructure.

Follow more of the Liftoff with Keith:

- Spotify: https://open.spotify.com/show/3cFpLXfYvcUsxvsT9MwyAD

- Apple Podcasts: https://podcasts.apple.com/us/podcast/liftoff-with-keith-newman/id1560219589

- Substack: https://keithnewman.substack.com/

- LinkedIn: https://www.linkedin.com/company/liftoffwithkeith/

- Newman Media Studios: https://newmanmediastudios.com/

For sponsorship inquiries, please contact: sponsorships@wherewithstudio.com

Attachment theories with special guest Jim Farm - Part 1