The race to a million qubits

Overview

Quantum computers promise to revolutionize drug discovery, cryptography, and materials science—but only if they can scale from today's dozens of qubits to millions. The bottleneck isn't the qubits themselves. It's the massive control systems: each qubit needs laser beams precisely controlled by bulky modulators, and thousands of cables snaking from room temperature into refrigerators colder than deep space. A lab rack that controls 100 qubits would need an entire data center to control a million.

Now the semiconductor industry is adapting the trick that powered the digital revolution: shrink the control hardware using the same CMOS chip factories that make your phone's processor. CU Boulder's December 2025 breakthrough—an optical modulator 100 times thinner than a hair, consuming 80 times less power—proves quantum control can be mass-produced. Meanwhile, Silicon Quantum Computing shattered accuracy records with 99.99% gate fidelity in silicon qubits that actually improve as they scale. SEALSQ, PsiQuantum, IBM, and Google are converging on CMOS manufacturing as the path to million-qubit machines by 2027-2030, with over $1 billion in fresh funding and multiple construction sites now breaking ground.

Play on this story Voices Debate Predict

Key Indicators

99.99%

Silicon qubit fidelity record

Silicon Quantum Computing's gate fidelity—qubits improve as they scale

$1B

PsiQuantum Series E funding

Raised September 2025 to build million-qubit photonic systems

2027

PsiQuantum million-qubit target

Most aggressive timeline in the industry, construction underway

4,158

Current qubit record

IBM's connected Kookaburra chips demonstrated in 2025

Interactive

Exploring all sides of a story is often best achieved with Play.

Ever wondered what historical figures would say about today's headlines?

People Involved

Matt Eichenfield

Leading CMOS photonics research for quantum control systems

Jake Freedman

Lead researcher on CMOS optical modulator breakthrough

Nils Otterstrom

Co-senior author on CU Boulder collaboration, leading EPIQ program

Carlos Moreira

Leading commercialization of CMOS-based quantum computing

Organizations Involved

University of Colorado Boulder - Department of Electrical, Computer & Energy Engineering

University Research Department

Leading CMOS photonics research for quantum computing

Academic department developing piezoelectric optomechanical photonic circuits for quantum computers.

Sandia National Laboratories

Federal Research Laboratory

Collaborating on CMOS photonics for quantum computing

DOE laboratory developing integrated photonics platforms with $17M EPIQ program funding.

IBM Quantum

Corporate research and commercial division

Targeting million-qubit systems by 2030

Leading quantum computing developer with aggressive scaling roadmap toward fault-tolerant systems.

Google Quantum AI

Corporate Research Lab

Developing error-corrected systems toward million-qubit goal

Achieved exponential error reduction milestone with Willow chip in late 2024.

PsiQuantum

Private Quantum Computing Company

Construction underway on Chicago and Brisbane facilities for million-qubit systems by 2027

Photonic quantum computing company with aggressive commercialization timeline.

SEALSQ Corp

Semiconductor & Quantum Computing Company

Commercializing CMOS-based quantum computing 2026-2030

Publicly-traded company (NASDAQ: LAES) developing silicon-based quantum computing using existing CMOS manufacturing infrastructure.

Silicon Quantum Computing (SQC)

Private Quantum Computing Company

Achieved record 99.99% gate fidelity with silicon qubits in December 2025

Australian company developing silicon-based quantum processors with atom-level precision manufacturing.

Timeline

Russia Demonstrates 72-Qubit Neutral Atom System
Technical Milestone

Lomonosov Moscow State University tested 72-qubit neutral rubidium atom quantum computer with 94% two-qubit operation accuracy.
December 27th, 2025
CU Boulder Unveils CMOS Quantum Control Chip
Manufacturing Breakthrough

Nature Communications published CU Boulder's CMOS-fabricated optical modulator 100x smaller than hair, using 80x less power.
December 26th, 2025
SEALSQ Announces CMOS Quantum Manufacturing Strategy
Strategic Announcement

SEALSQ unveiled 2026-2030 plan to commercialize silicon-based quantum computing using existing CMOS semiconductor foundries and supply chains.
December 19th, 2025
Silicon Quantum Computing Achieves Record 99.99% Gate Fidelity
Manufacturing Breakthrough

Nature published SQC's silicon-based multi-register processor with 99.99% gate fidelity using CMOS-compatible manufacturing—qubits improve accuracy as they scale.
December 17th, 2025
IBM Demonstrates 4,158-Qubit System
Technical Milestone

IBM connected three Kookaburra chips into 4,158-qubit system and unveiled 120-qubit Nighthawk processor.
November 12th, 2025
Google Demonstrates Verifiable Quantum Advantage with Willow
Technical Milestone

Google's Quantum Echoes algorithm on 105-qubit Willow chip ran 13,000 times faster than classical supercomputers, marking first verifiable quantum advantage.
October 22nd, 2025
PsiQuantum Raises $1 Billion for Million-Qubit Systems
Funding Announcement

PsiQuantum closed $1 billion Series E funding at $7 billion valuation to build fault-tolerant photonic quantum computers, with construction sites planned for Brisbane and Chicago.
September 10th, 2025
IBM Breaks Ground on Fault-Tolerant System
Strategic Announcement

IBM announced world's first large-scale fault-tolerant quantum computer construction in Poughkeepsie data center.
June 15th, 2025
Google's Willow Crosses Error Threshold
Technical Milestone

Google's 105-qubit Willow chip achieved exponential error reduction with increasing qubits, proving error correction scales.
December 9th, 2024
Sandia Labs Launches $17M EPIQ Program
Funding Announcement

Sandia awarded $17 million for Error-Corrected Photonic Integrated Qubits Grand Challenge program.
August 22nd, 2024
IBM Delivers 1,121-Qubit Condor
Technical Milestone

IBM unveiled Condor processor with 1,121 superconducting qubits, exceeding 1,000-qubit milestone.
December 4th, 2023
Intel Debuts Cryogenic Control Chip
Hardware Innovation

Intel introduced Horse Ridge II, cryogenic control chip addressing quantum computing's wiring bottleneck.
April 28th, 2021
IBM Unveils Quantum Roadmap
Strategic Announcement

IBM announced aggressive scaling roadmap targeting 1,000+ qubits by 2023 and ultimately millions of qubits.
September 15th, 2020
Google Claims Quantum Supremacy
Technical Milestone

Google's 53-qubit Sycamore processor performed calculation in 200 seconds that would take classical supercomputer 10,000 years.
October 23rd, 2019

Scenarios

Predict which scenario wins. Contrarian picks score more — points lock in when the scenario resolves.

Predictions leaderboard

CMOS Manufacturing Unlocks Million-Qubit Systems by 2030

CMOS-fabricated control hardware solves the scaling bottleneck. Semiconductor foundries begin mass-producing integrated photonic modulators, cryogenic controllers, and qubit chips using proven 300mm wafer processes. By 2029-2030, IBM delivers its Quantum Starling system with 200 logical qubits performing 100 million error-corrected operations, while PsiQuantum launches its million-qubit photonic system. Early commercial applications emerge in drug discovery and materials science. The quantum computing market expands from $3.5 billion in 2025 to over $20 billion by 2030 as practical advantage becomes demonstrable.

Discussed by: IBM roadmap projections, PsiQuantum timeline, industry analysts at Bain & Company

Consensus —

Scaling Stalls at Thousands of Qubits Through 2030s

CMOS manufacturing helps but doesn't solve all bottlenecks. Error rates remain too high, even with advanced error correction consuming most qubits for redundancy. Power consumption for cryogenic cooling scales super-linearly, making million-qubit systems economically impractical. Quantum systems plateau around 10,000-50,000 physical qubits through the early 2030s, delivering only hundreds of logical qubits. The technology finds niche applications but commercial viability remains a decade away, validating predictions that practical quantum computing requires a million-fold improvement and won't arrive until the 2040s.

Discussed by: Nvidia CEO Jensen Huang, conservative quantum researchers, cited in Financial Times coverage

Consensus —

Alternative Architecture Leapfrogs Superconducting Approach

Room-temperature or higher-temperature quantum computing approaches mature faster than expected. Photonic systems eliminate cryogenic requirements entirely. Silicon spin qubits achieve integration density that superconducting qubits cannot match. Novel architectures like QuamCore's patented design integrate a million qubits in a single cryostat—previously thought impossible. By 2030, the winning approach isn't superconducting qubits with CMOS control, but a fundamentally different architecture. IBM and Google's massive investments in superconducting systems become stranded assets as the industry pivots, echoing past technology transitions.

Discussed by: Photonic quantum computing advocates, silicon spin qubit researchers, emerging startups like QuamCore

Consensus —

Historical Context

Transistor Integration and Moore's Law (1960-2000)

1960-2000

What Happened

The semiconductor industry faced repeated scaling crises as transistor counts climbed from hundreds to billions. Each time, engineers predicted physics would halt progress. Each time, manufacturing innovations—photolithography advances, new materials, 3D architectures—extended Moore's Law. The turning point came when the industry standardized CMOS fabrication in the 1980s, enabling massive economies of scale. Companies built multi-billion dollar foundries producing millions of identical chips, driving exponential cost reductions per transistor.

Outcome

Short Term

By 1990, CMOS manufacturing enabled microprocessors with millions of transistors at consumer prices.

Long Term

CMOS became the foundation of the digital revolution, with foundries producing billions of chips annually by the 2000s.

Why It's Relevant Today

Quantum computing faces an identical challenge: control systems don't scale. The CU Boulder breakthrough follows the exact playbook—adapt CMOS foundries to mass-produce quantum hardware, leveraging trillions in semiconductor investment.

IBM Mainframe Crisis and the PC Revolution (1980s)

1980-1990

What Happened

IBM dominated computing with room-sized mainframes requiring specialized cooling and power infrastructure. When personal computers emerged using standardized components and simpler manufacturing, skeptics argued they'd never match mainframe capabilities. But standardization and mass production drove exponential cost-performance improvements. By the late 1980s, desktop machines matched mainframe power at a fraction of the cost and complexity.

Outcome

Short Term

PC sales exploded in the mid-1980s as price-performance crossed critical thresholds for businesses and consumers.

Long Term

Mainframes became niche products while distributed computing on standardized hardware became the paradigm.

Why It's Relevant Today

Today's quantum computers resemble 1970s mainframes: refrigerator-sized systems with massive infrastructure. If CMOS manufacturing enables compact, standardized quantum processors, the disruption could mirror the PC revolution.

Fiber Optics Supplants Copper Networks (1990-2010)

1990-2010

What Happened

Telecommunications faced a bandwidth crisis in the 1990s as internet traffic exploded. Copper networks couldn't scale. The industry bet on fiber optics, but early systems were expensive and difficult to manufacture. Breakthroughs in optical component manufacturing—particularly CMOS-compatible photonic chips in the 2000s—enabled mass production. Optical transceivers that cost thousands of dollars dropped to tens of dollars.

Outcome

Short Term

By 2000, fiber backbone networks began replacing copper for long-haul internet traffic.

Long Term

By 2010, fiber reached homes and data centers globally; photonic chips became commodities.

Why It's Relevant Today

CU Boulder's optical modulator uses the same CMOS-photonics integration that enabled cheap fiber optics. If quantum control optics follow the same cost curve, million-qubit systems become economically feasible.