Blowing Past the AI Scaling Wall

Probabilistic Computing with Magnetic Tunnel Junctions and Digital CMOS

What if your computer didn't need to be certain to be powerful? In this episode, we take a deep dive into Professor Pedram Khalili's work in the emerging field of probabilistic computing — a fundamentally different approach to tackling problems that have stumped classical systems for decades. From cracking integer factorization to navigating complex energy landscapes, Khalili shows how pairing magnetic tunnel junctions with CMOS circuits unlocks speed and efficiency that deterministic computing simply can't match. Along the way, he introduces p-dits, a multi-dimensional twist on traditional computing variables, and makes the case for a future where cutting-edge spintronic devices and scalable digital architectures finally speak the same language. Essential listening for anyone curious about the hardware frontier reshaping AI and beyond.

Beyond CMOS: Ballistic Fluxons and the Future of Reversible Computing with Dr. Kevin Osborn, University of Maryland

What if computers could think faster while using a fraction of the energy? That future may already be in the lab. In this episode of NSF Discover Superconducting, we take a deep dive into a talk by Dr. Kevin Osborn of the Joint Quantum Institute at the University of Maryland — exploring ballistic and reversible superconducting logic, a radical rethinking of how digital gates work at the quantum level. As AI data centers push toward 20–30 megawatt power loads, Osborn's research couldn't be more timely. His team is building logic gates that harness the momentum of fluxons — magnetic flux quanta — to process information without constant power input. The result? Simulations showing over 97% energy efficiency and a potential path to zeptojoule-level computing. We cover the physics behind Long Josephson Junctions, the Ballistic Flip-Flop (BFF), two-polarity bit systems using fluxons and anti-fluxons, and a theoretical framework for sub-nanosecond qubit readout. Plus: early experimental results from the MIT Lincoln Labs fabrication process. Perfect for students and researchers in quantum computing, electrical engineering, and sustainable technology.

What comes after CMOS?

In this episode, we explore the future of superconductor electronics—a promising post-CMOS computing paradigm offering ultra-low energy consumption and ultra-high processing speeds. Dr. Sasan Razmkhah of USC’s SPORT Lab joins the conversation to discuss cutting-edge research aimed at making superconductor very-large-scale integration (sVLSI) a practical reality. We dive into the development of Fast-Phase Logic (FPL), a next-generation logic family that uses π-Josephson junctions and stacked zero-Josephson junctions to dramatically increase logic density while reducing power requirements. The discussion also covers hybrid system architectures that integrate superconductor logic with CMOS, spanning multiple temperature zones to balance performance, efficiency, and manufacturability. Together, these advances outline a roadmap toward scalable, energy-efficient computing systems—pointing to a future where superconductors play a central role in high-performance and AI-driven computing.

Blowing Past the AI Scaling Wall

This deep-dive episode explores the ideas from Jeff Shainline’s recent seminar on superconducting optoelectronic networks (SOENs) and the radical rethink of AI hardware they represent. We unpack how SOENs combine analog superconducting circuits, integrated memory, and single-photon optical communication to bypass the von Neumann bottleneck and deliver massive gains in speed, energy efficiency, and scalability. If you’re curious about brain-inspired hardware, trillion-parameter AI, or the future of computation, this episode guides you through the key concepts and why they matter.

Quantum Signals & Superconductivity: Insights from Dr. Sam Benz at NIST

In this episode, we recap highlights from a seminar by Dr. Sam Benz of the NIST Superconductive Electronics Group, exploring his team’s cutting-edge work in quantum-based signal synthesis and precision measurement systems. Dr. Benz’s research focuses on superconducting voltage sources, quantum reference systems, and ultra-precise tools used by advanced laboratories and industries around the world. We break down key takeaways from his talk, including developments in high-speed computing, quantum technologies, and RF communications — all made possible by superconductive electronics. Whether you're new to the topic or looking to better understand where quantum technology is headed, this seminar recap makes the science both clear and compelling. #QuantumSignals #Superconductivity #NIST #EngineeringPodcast #ScienceExplained #STEM