Diraq and Dell Test Hybrid Quantum-Classical Computing

Diraq and Dell Technologies installed a Dell high-performance computing cluster inside Diraq's Sydney laboratory on Sept. 18, physically next to Diraq's silicon spin-qubit quantum processor, the companies said. It is a small deployment: a single cluster, not a production data center, built to test how classical and quantum hardware talk to each other when the network distance between them is measured in meters rather than continents.

What co-location is supposed to buy

Diraq designs quantum processors using silicon spin qubits fabricated on standard CMOS manufacturing lines, an approach the Sydney-founded company says can eventually pack millions of qubits onto a single chip. Andrew Dzurak, Diraq's founder and chief executive, said the goal of the Dell partnership is a system where ordinary workloads stay on classical hardware and only the hard parts get handed off. "We're working with Dell to create integrated systems that call classical AI and CPUs for the big number crunching," Dzurak said, according to Diraq. "Then, when there's a specific hard problem, you call the quantum processing unit."

John Roese, Dell's global chief technology officer and chief AI officer, framed it as an industry-wide shift: "The term of the day is hybrid quantum-classical computing," Roese said. "The really complex problems in the world need systems that bring different types of compute together." The initial work covers connectivity validation and latency testing, moving toward automated qubit calibration, tuning and early error-correction routines, according to The Quantum Insider, which reported the deployment two days after Diraq's own announcement. Target applications named by the companies include drug discovery, logistics optimization and financial modeling, none of which the companies said they have yet run on the system. The near-term engineering work is less glamorous than any of those applications: adapting Dell's orchestration software to automate real-time qubit calibration and tuning, and to carry early quantum error-correction routines, the kind of housekeeping that has to work before a hybrid system can be trusted with a real workload.

The funding trail behind the announcement

Diraq spun out of the University of New South Wales in 2022 and has since raised more than $150 million, including government money, StartupDaily reported on Feb. 3, 2026, when Australia's National Reconstruction Fund Corporation put in 20 million Australian dollars as part of an ongoing 75 million Australian dollar Series A round. Dzurak connected that raise directly to the kind of integration Dell is now testing, saying the investment arrives as "Australia builds its strength in critical technology infrastructure, particularly within our booming data centre sector," and that Diraq's processors are "natively designed to integrate seamlessly with existing data centres."

Date	Development
Feb. 3, 2026	NRFC invests AU$20 million in an ongoing AU$75 million Series A; total funding tops $150 million
May 21, 2026	Diraq signs a $38 million letter of intent under the U.S. CHIPS Act with GlobalFoundries
Sept. 18, 2026	Dell co-locates an HPC cluster inside Diraq's Sydney lab

Three months after the NRFC round, Diraq signed a $38 million letter of intent with the U.S. Department of Commerce's CHIPS Research and Development Office, partnering with chipmaker GlobalFoundries, The Quantum Insider reported on May 21. That figure is a letter of intent, not committed funding, a preliminary step that still requires further government action before any money moves. GlobalFoundries chief technology officer Gregg Bartlett said the company is "proud to partner with Diraq to advance silicon-based quantum processors," tying Diraq's chip design to the same CMOS fabrication lines that make conventional processors, William Jeffrey, chairman of Diraq's board and a former director of the U.S. National Institute of Standards and Technology, framed the CHIPS Act money in national-security terms rather than commercial ones: "Quantum technology is a matter of national competitiveness," Jeffrey said. That framing sits oddly next to a partnership whose own numbers are limited to funding totals and a room-sharing arrangement; the government money is committed to the company's manufacturing roadmap months before Dell's own testbed has produced a single public benchmark of what that roadmap is supposed to deliver.

What neither company said

Neither Diraq nor Dell published a measured latency figure for the new link, a qubit count for the processor sitting next to Dell's cluster, or a benchmark showing the hybrid setup outperforming a classical-only run on any task. The Quantum Insider's own write-up flagged the absence directly, describing Diraq's announcement as carrying "no independent verification or analysis from external sources." That gap matters more here than in a typical hardware launch: the entire pitch for co-location is that physical proximity cuts communication latency enough to matter, and a claim about latency with no latency number attached is a claim about geography, not performance.

What would confirm the pitch is a published number: a round-trip time between the Dell cluster and the Diraq processor, in microseconds or better, run against a disclosed baseline for a non-co-located system. Until Diraq or Dell publishes one, the Sydney testbed demonstrates that a server and a quantum chip can share a room, which is a real engineering step but a smaller one than the announcement's framing suggests. The same discipline applies to compute infrastructure elsewhere on this beat: System76's Thelio Mira AI workstation shipped five GPU memory tiers with no benchmark numbers attached either, and Intel's Crescent Island inference chip launched around a tokens-per-watt pitch minus the tokens-per-watt figure. A number is the difference between a specification and a claim.