Microsoft Opens New Quantum Research Center in Maryland, Accessible to DARPA
Microsoft has opened a new quantum research center in Maryland, allowing the U.S. Department of Defense's Defense Advanced Research Projects Agency (DARPA) to operate and test its next-generation quantum systems on-site.
The research center is located in the University of Maryland's Discovery District, near Washington, D.C., covering approximately 15,000 square feet and set to be operational by September 22, 2026. The quantum systems deployed at the center utilize Microsoft's second-generation topological quantum chip, Majorana 2, which replaces aluminum with lead in its material structure to enhance performance. The facility also features a quantum hardware maker space, hosts measurement-based quantum computing workshops annually, and funds postdoctoral researchers at the University of Maryland.
Previously, DARPA could only remotely assess Microsoft's quantum technology, with equipment located at the headquarters in Redmond, Washington, and in Europe. Now, DARPA can fully access the site for independent testing. Zulfi Alam, Vice President of Microsoft's Quantum Business, stated that this "means DARPA can personally verify the equipment," assessing the machine's true performance; such external evaluations will "enforce a level of engineering rigor that research teams typically lack." He also referred to DARPA as a "completely trustworthy institution."
Microsoft is participating in DARPA's "Quantum Benchmarking Initiative" (QBI) under the "Underexplored Systems for Practical Scale Quantum Computing" (US2QC) project, which has entered the final evaluation phase. DARPA is simultaneously evaluating multiple quantum companies to determine which technological routes can produce commercially viable and economically feasible quantum computers by 2033. Microsoft's own goal is to launch a commercially viable quantum system by 2029.
In terms of technical collaboration, Microsoft's quantum ecosystem partners include AMD, Intel, cryogenic equipment supplier Bluefors, Finnish quantum company IQM, the U.S. Fermi National Accelerator Laboratory, quantum error correction software company Riverlane, and Quantum Motion. Additionally, Microsoft is collaborating with neutral atom quantum company Atom Computing to develop a quantum computer named "Magne," designed to have 50 logical qubits, which is planned to be operational in early 2027 at QuNorth in Denmark. This indicates that while Microsoft is developing its topological route, it is also betting on the neutral atom route.
Quantum computing's core significance for national security lies in encrypted communication: it can be used both to protect encrypted communications and to break existing encryption systems. Deploying the system near Washington and allowing DARPA to conduct on-site testing signifies that Microsoft's quantum project is transitioning from the corporate R&D phase into the independent verification process of the U.S. defense system.
From a market mechanism perspective, this is a contract and policy-driven event, rather than a transaction-driven one. DARPA's evaluation results will effectively determine which quantum technology routes can receive endorsement and subsequent procurement budgets from the U.S. government, impacting capital allocation across the entire quantum sector. Beneficiaries include Microsoft, which has entered the final evaluation phase, as well as supply chain partners in cryogenic equipment, control chips, and error correction software. Companies directly competing with Microsoft on the technology front, including public quantum firms like IonQ, Rigetti, and D-Wave, as well as IBM and Google, which are pursuing superconducting routes, will face pressure as they need to demonstrate their practical progress under the same government benchmarks.
Source: Public Information
ABAB AI Insight
Microsoft's bet on the topological quantum route is one of the most controversial in the industry. In 2018, a team from Delft University of Technology, funded by Microsoft, published a paper in Nature claiming to have observed the quantum conductance of Majorana zero modes, but it was retracted in 2021 due to data processing issues. In February 2025, Microsoft released its first-generation chip, Majorana 1, claiming to have achieved 8 topological qubits using a "topological conductor" made of indium arsenide and aluminum, but the accompanying paper published in Nature did not prove the actual existence of the topological state, leading several physicists to publicly question its "topological gap protocol" testing method. By handing the system over to DARPA for on-site testing, Microsoft is, in a sense, responding to external criticisms of "only looking at papers, unable to verify."
From a capital path perspective, Microsoft is adopting a strategy of "mainline self-research plus multiple route hedging." It has invested nearly 20 years in the topological quantum direction while collaborating with Quantinuum in 2024 to demonstrate logical qubits encoded by physical qubits. In 2025, its collaboration with Atom Computing on the Magne project will take place in Denmark, supported by the Novo Nordisk Foundation and the Danish Export and Investment Fund (EIFO), with QuNorth investing approximately 80 million euros. In the cloud, Azure Quantum has integrated with several third-party hardware providers, including IonQ, Quantinuum, and Rigetti. This way, regardless of which technology route ultimately prevails, Microsoft can maintain its foothold in quantum computing distribution through Azure. Establishing the research center near Washington is aimed at aligning with the procurement and testing needs of defense and intelligence systems.
Historically, this model is reminiscent of DARPA's funding of ARPANET (the precursor to the internet) in the 1960s and the 2004 Grand Challenge for autonomous vehicles: defense agencies set objective benchmarks, filter technology routes, and then hand the winning routes over to industry for commercialization. In the quantum sector, Google plans to release its Willow chip with 105 physical qubits in December 2024, demonstrating "below threshold" error correction for the first time; IBM will unveil its Starling roadmap in June 2025, aiming for 200 logical qubits by 2029. Quantinuum is set to raise $600 million at a valuation of about $10 billion in 2025, while IonQ will acquire Oxford Ionics for approximately $1.075 billion. The entire industry is transitioning from a "competition of physical qubits to logical qubits and engineering validation," with DARPA's independent benchmark testing serving as a form of elimination round.
Essentially, this represents a regulatory change; more accurately, the state has taken control of the pricing power of technology routes through benchmark testing. Quantum computing has long lacked unified, credible performance standards, with companies making various claims regarding the number of physical qubits and "quantum supremacy," leaving capital to rely on companies' own marketing for pricing. DARPA's QBI sets the threshold of "economically feasible" by 2033 and insists on independent verification, shifting the evaluation standard from self-certification by companies to third-party certification by the government. The driving force behind this is cryptographic security: the National Institute of Standards and Technology (NIST) is set to release the first post-quantum cryptography standards in 2024, and the National Security Agency (NSA) has mandated the transition of national security systems by 2030 to 2035, as adversaries can "capture now and decrypt later." Once quantum computing is defined as a national security capability, R&D funding, procurement budgets, and technology route choices will concentrate on a few companies validated by the government.
ABAB News · Cognitive Law
Who sets the testing standards decides the winning and losing of technology routes.
Papers persuade peers, while actual tests persuade budgets.
Bet on a mainline while keeping all route entrances secure.