632nm

Building Quantum Computers at Semiconductor Scale | John Martinis on Qolab + Superconducting Qubits

26 snips
Sep 22, 2026
John Martinis, physicist and superconducting quantum-computing pioneer, discusses what it takes to build machines with millions of qubits. Topics include tunable couplers, coherence, wiring, cryogenic power, material defects, error correction, and semiconductor-scale fabrication. He also reflects on Google’s Sycamore experiment, lessons from failure, and Qolab’s vision for dense, modular quantum hardware.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Why A Million Qubits Breaks Conventional Control

  • Scaling superconducting systems requires solving wiring and cryogenic power, not merely duplicating qubit cells.
  • Ten milliwatts per cryogenic control chip would become roughly 10 kilowatts across a million-qubit machine.
ADVICE

Use Motion To Suppress Qubit Drift

  • Randomize errors with spin echoes and gauge transformations instead of relying solely on perfectly stable qubit frequencies.
  • Martinis says the worst operation for superconducting qubits can be doing nothing, because drift accumulates unchecked.
INSIGHT

Tiny Surface Defects Become A Scaling Crisis

  • Two-level states arise from microscopic defects in insulating surfaces and can resonantly steal qubit energy.
  • They are rare enough to avoid through tuning, but their growing number makes that strategy painful at scale.
Get the Snipd Podcast app to discover more snips from this episode
Get the app