I have been thinking for the past few weeks about where quantum computer chip fabrication actually happens in the United States. By quantum fabs, I mean facilities capable of fabricating quantum devices, not mass production CMOS plants. What emerges instead is a small set of research oriented or pilot scale fabs designed to iterate on fragile devices rather than produce them at volume.
You know I have been really digging into Rigetti over the last few weeks. Rigetti’s long term competitive position may hinge less on near term qubit counts and more on whether it successfully internalizes quantum chip fabrication as a core capability. As the note from the Nasdaq analysis team notes, Rigetti’s pursuit of an in-house quantum fab reflects a strategic bet that control over materials, process integration, and yield learning could become a defensible moat in superconducting quantum computing [1]. Owning fabrication allows tighter iteration between device physics and system architecture, potentially accelerating coherence improvements and reducing dependency on external foundries that are optimized for classical CMOS. At the same time, this strategy materially raises capital requirements and execution risk in an industry already defined by low revenue and long timelines. If Rigetti can translate fab control into demonstrably superior device performance and reliability, the investment could differentiate it from peers. If not, the fab risks becoming a costly fixed asset in a market that has yet to prove durable commercial demand.
During its third quarter earnings call, the Rigetti Computing (RGTI) team acknowledged that advancing beyond its current development trajectory will ultimately require more sophisticated fabrication infrastructure. Management emphasized that the company’s existing 150 mm Fremont facility is sufficient to support planned progress through roughly 2027, but it is unlikely to deliver the greater than 99.9% gate fidelities associated with commercially viable quantum systems. Rather than treating this limitation as a distant or abstract risk, Rigetti has articulated a set of concrete pathways forward, including potential partnerships with superconducting focused foundries, participation in emerging United States government quantum manufacturing programs, and contingency planning for a next generation 200 mm to 300 mm fabrication facility with higher levels of automation and process control. This is really more or less a real geopolitical risk in terms of where fabs are going to be located.
We really do want to support building these things in the United States from the start. All of the backward linkages to other technology within the ecosystem is dependent on having these innovation hubs. This got me thinking about how many quantum computer chip fabs we actually have in the United States. My searches have yielded 7 results:
Rigetti Computing (Fremont, California): 150 mm captive superconducting fab [1].
IBM Thomas J. Watson Research Center (Yorktown Heights, New York): 200 mm research scale quantum chip fab [2].
IBM Albany NanoTech Complex (Albany, New York): 300 mm advanced fab used for IBM quantum processor fabrication [2][3].
Google Quantum AI (Santa Barbara, California): quantum fabrication facility on campus, wafer size not stated [4].
Sandia National Laboratories (Albuquerque, New Mexico): MESA microsystems fabrication complex with CMOS, silicon photonics, and III-V processing capabilities [5][6].
Intel (Hillsboro, Oregon): D1 facility used to fabricate silicon spin qubits on 300 mm wafers [7][8].
PsiQuantum (Chicago, Illinois): IQMP buildout and deployment site, chips fabricated via foundry partners, avoid calling it a 300 mm photonics foundry [9][11].
I actually thought we would see more of these fabs, but maybe the cost of setting one of these up is just higher than people are willing to invest. My guess here is that more research level fabs may exist and I’m going to need to dive into these more going forward.
Building a quantum computer fabrication facility requires a blend of semiconductor-grade manufacturing tools, precision cryogenic and photonics systems, and specialized qubit-dependent equipment. A true quantum chip fab is closer to a research foundry than a high-volume CMOS fab, yet it still needs many of the same multi-million-dollar systems used in advanced microelectronics.
Let’s break that investment down into 9 parts. In future posts, I plan on deep diving each of these and who might be the major players in each category in more detail.
Cleanroom Infrastructure & Environmental Isolation
Advanced Lithography Systems
Precision Deposition Tools
Etching & Material Removal
Metrology & In-Line Inspection
Wafer Packaging & 3D Integration
Cryogenic Test Infrastructure
Photonics & Optical Control Systems
Supporting Utilities & Safety Systems
A quantum computer fabrication facility is inherently a hybrid environment that blends classical semiconductor manufacturing equipment with low temperature and microwave engineering infrastructure, photonics and vacuum engineering tools, and highly controlled cleanroom and material processing systems. This convergence reflects the reality that quantum hardware inherits its physical foundations from microelectronics while introducing entirely new constraints driven by coherence, noise suppression, and cryogenic operation. As a result, capital requirements scale quickly. Research-scale facilities can be established with investments in the tens of millions of dollars, while production-grade fabs capable of repeatable device fabrication, integrated testing, and early manufacturing scale routinely push into the hundreds of millions as cleanroom complexity, tool depth, and cryogenic test capacity expand.
Things to consider:
Fabrication control may become a stronger differentiator than raw qubit counts.
The scarcity of quantum fabs highlights manufacturing, not algorithms, as a near term bottleneck.
Government backed manufacturing programs may shape which architectures survive.
Capital intensity favors firms with long time horizons and patient funding.
Scaling quantum hardware looks more like advanced materials science than software iteration.
Footnotes:
[1] Gupta, H. (2025, December 10). Can Rigetti’s need for a quantum fab reshape its long-term moat? Nasdaq. https://www.nasdaq.com/articles/can-rigettis-need-quantum-fab-reshape-its-long-term-moat
[2] IBM Quantum, “Building quantum computers with advanced semiconductor technology”, https://www.ibm.com/quantum/blog/300mm-fab
[3] NY CREATES, “IBM Announces All Future Chips on the IBM Quantum Development Roadmap to be Fabricated at NY CREATES Albany NanoTech Complex”, https://ny-creates.org/ibm-announces-all-future-chips-on-the-ibm-quantum-development-roadmap-to-be-fabricated-at-ny-creates-albany-nanotech-complex/
[4] Google Quantum AI, “Our lab”, https://quantumai.google/lab
[5] Sandia National Laboratories, “Microsystems Engineering, Science and Applications (MESA)”, https://www.sandia.gov/mesa/
[6] Sandia National Laboratories, “MESA capabilities overview”, https://www.sandia.gov/mesa/capabilities-overview/
[7] Intel, “Intel’s New Chip to Advance Silicon Spin Qubit Research for Quantum Computing”, https://www.intc.com/news-events/press-releases/detail/1626/intels-new-chip-to-advance-silicon-spin-qubit-research
[8] Nature, “Probing single electrons across 300-mm spin qubit wafers”, https://www.nature.com/articles/s41586-024-07275-6
[9] PsiQuantum, “PsiQuantum Breaks Ground on America’s Largest Quantum Computing Project in Chicago”, https://www.psiquantum.com/news-import/psiquantum-breaks-ground-chicago
[10] University of Chicago News, “Groundbreaking of Illinois Quantum and Microelectronics Park creates anchor for quantum innovation”, https://news.uchicago.edu/story/groundbreaking-illinois-quantum-and-microelectronics-park-creates-anchor-quantum-innovation
[11] Reuters, “GlobalFoundries buys Singapore’s Advanced Micro Foundry in push to speed up AI data center networks”, https://www.reuters.com/world/asia-pacific/globalfoundries-buys-singapores-advanced-micro-foundry-push-speed-up-ai-data-2025-11-18/
