Microsoft is giving independent evaluators a closer look at its quantum hardware.
Microsoft opened a 15,000-square-foot quantum research facility in the University of Maryland’s Discovery District on Tuesday, granting the Defense Advanced Research Projects Agency direct, physical access to its proprietary quantum hardware.
While DARPA previously evaluated Microsoft’s systems remotely across locations in Redmond, Washington, and Europe, evaluators now have a dedicated workspace where they can load hardware and run the system through their own boot sequences.
DARPA will initially test a system using Microsoft’s Majorana 2 chip. Microsoft says its topological approach could reduce the resources needed for error correction, but independent testing has yet to establish how the system performs.
"This is a big deal for us because it essentially enables DARPA to kick the tires and to sort of figure out how good this machine is and how it's performing," Zulfi Alam, corporate vice president of Microsoft Quantum, told Reuters. Alam added that the evaluation "enforces a certain amount of engineering rigor that scientific teams normally do not have."
The road to utility-scale compute
Microsoft had already reached the final phase of DARPA’s Underexplored Systems for Utility-Scale Quantum Computing program. The Maryland facility now gives evaluators on-site access during that phase, a component of the broader Quantum Benchmarking Initiative. DARPA is screening multiple commercial quantum architectures to establish whether any approach can deliver an economically viable, utility-scale machine by 2033. Microsoft is aiming for an even more aggressive timeline, targeting commercial systems by 2029.
To verify whether Microsoft's topological architecture holds up, DARPA’s evaluation unites specialists from the Air Force Research Laboratory, the Johns Hopkins University Applied Physics Laboratory, and four national laboratories: Los Alamos, Oak Ridge, Lawrence Berkeley, and Lawrence Livermore.
Beyond defense testing, the facility includes an educational quantum makerspace backed by Intel, AMD, Riverlane, Fermilab, and IQM, designed to train researchers on hands-on hardware assembly.
The litmus test for topological qubits
By inviting federal defense researchers to directly probe its hardware, Microsoft is taking a calculated reputational gamble to resolve years of scientific debate. The tech giant's topological architecture has long faced skepticism across the quantum physics community over whether its underlying quasiparticle phenomena can reliably produce functional qubits.
Unlike peer rivals IBM and Google, which have focused on superconducting circuits, Microsoft staked its long-term roadmap on an exotic, high-risk topological framework. Giving DARPA physical access puts Microsoft’s engineering claims under closer independent scrutiny.
DARPA’s evaluation could show which parts of Microsoft’s approach work as intended and which still need development. Its findings would offer a more useful measure of progress than a chip announcement alone, though results have not yet been reported.
What eWeek found: Enterprise takeaways
Independent government benchmarking alters how IT decision-makers and enterprise leaders should budget for quantum computing over the next five years:
Procurement reality checks: DARPA’s primary metric measures whether the economic value of a quantum calculation outweighs the system's operational cost. Enterprise leaders evaluating post-classical compute should adopt this identical benchmark rather than tracking raw physical qubit counts.
Dual-track architecture strategy: Microsoft is insulating its commercial software stack against potential topological hardware delays. Through partnerships like Atom Computing’s 50-logical-qubit neutral-atom system "Magne," scheduled for 2027, Microsoft ensures Azure Quantum customers can build fault-tolerant applications regardless of when Majorana chips hit production.
Long-term encryption timelines: DARPA’s active hardware stress-testing indicates that cryptographic vulnerabilities tied to quantum decryption remain an active national security concern, reaffirming that enterprises should continue prioritizing post-quantum cryptography migration across existing data infrastructure.
Read more: IBM researchers reported cutting the number of samples needed for one quantum experiment by up to 63-fold, another step toward making today’s hardware more useful.