Optimizing Google’s AI infrastructure: Albert Cohen’s (Google DeepMind) keynote is now online!
At SCAI's annual "Warm Up" event on the Sorbonne University campus, Albert Cohen, Research Scientist at Google DeepMind, took the stage to reveal how AI is revolutionizing the design and optimization of tech infrastructure. The full video recording is now available.
Every year on the eve of VivaTech, SCAI brings together the AI ecosystem for its major kickoff evening. This year’s edition featured an exceptional keynote by Albert Cohen from Google DeepMind.
Rethinking the Synergy Between Hardware and Software
As AI models grow exponentially, the engineering and computing power challenges are massive. During his talk, Albert Cohen shared behind-the-scenes insights into how DeepMind applies machine learning to optimize and compile its own hardware architectures (such as Google's TPUs).
💡 Key Takeaways:
-
AI for Systems: How machine learning is used to optimize compilers and speed up raw computation.
-
The Scaling Challenge: Managing massive energy consumption and computing power requirements for next-gen models.
-
DeepMind's Playbook: Real-world feedback and collaborative research methods from one of the world's leading AI labs.
🎥 Watch the Full Keynote
Whether you are an engineer, researcher, or simply passionate about the future of tech, don't miss this deep dive.