Lex Fridman Podcast #225 – Jeffrey Shainline: Neuromorphic Computing and Optoelectronic Intelligence
126 snips
Sep 26, 2021 Jeffrey Shainline, a physicist at NIST, dives into the fascinating realm of neuromorphic computing and optoelectronic intelligence. He discusses how this innovative technology mimics brain functionality to revolutionize computing. The conversation covers superconductivity's role in enhancing computational speed and discusses the advantages of superconducting materials over traditional silicon. Shainline also explores the integration of light with electronics and challenges in maintaining ultra-low temperatures for advanced technologies. His insights promise to reshape our understanding of the future of computing.
AI Snips
Chapters
Transcript
Episode notes
Superconductivity Explained
- Superconductors, unlike semiconductors, allow for dissipationless current flow at low temperatures.
- This creates a macroscopic quantum state where electrons behave coherently.
Josephson Junctions
- Josephson junctions enable fast switching and signal propagation, potentially exceeding silicon's capabilities.
- However, scaling Josephson junctions remains a fundamental physical challenge.
Cooling Requirements
- Superconducting electronics require liquid helium cooling, making them impractical for personal devices.
- Their use is currently limited to large-scale systems like servers.