Host: Sebastian Hassinger
Guest: Andrew Dzurak (CEO, Diraq)
In this enlightening episode, Sebastian Hassinger interviews Professor Andrew Dzurak. Andrew is the CEO and co-founder of Diraq and concurrently a Scientia Professor in Quantum Engineering at UNSW Sydney, an ARC Laureate Fellow and a Member of the Executive Board of the Sydney Quantum Academy. Diraq is a quantum computing startup pioneering silicon spin qubits, based in Australia. The discussion delves into the technical foundations, manufacturing breakthroughs, scalability, and future roadmap of silicon-based quantum computers—all with an industrial and commercial focus.
Key Topics and Insights
1. What Sets Diraq Apart
- Diraq’s quantum computers use silicon spin qubits, differing from the industry’s more familiar modalities like superconducting, trapped ion, or neutral atom qubits.
- Their technology leverages quantum dots—tiny regions where electrons are trapped within modified silicon transistors. The quantum information is encoded in the spin direction of these trapped electrons—a method with roots stretching over two decades1.
2. Manufacturing & Scalability
- Diraq modifies standard CMOS transistors, making qubits that are tens of nanometers in size, compared to the much larger superconducting devices. This means millions of qubits can fit on a single chip.
- The company recently demonstrated high-fidelity qubit manufacturing on standard 300mm wafers at commercial foundries (GlobalFoundries, IMEC), matching or surpassing previous experimental results—all fidelity metrics above 99%.
3. Architectural Innovations
- Diraq’s chips integrate both quantum and conventional classical electronics side by side, using standard silicon design toolchains like Cadence. This enables leveraging existing chip design and manufacturing expertise, speeding progress towards scalable quantum chips.
- Movement of electrons (and thus qubits) across the chip uses CMOS bucket-brigade techniques, similar to charge-coupled devices. This means fast (<nanosecond scale) movement within the quantum processor, supporting complex quantum operations.
4. Cryogenic Operation
- Diraq’s qubits run at around 1 Kelvin, much warmer than superconducting qubits (which require millikelvin temperatures). This enables integration of classical CMOS control electronics at the same temperature layer, avoiding the wiring and cooling challenges typical in superconducting systems1.
5. Error Correction & Control
- Diraq aims for native error correction schemes adapted to their modular, but not fully 2D-grid, architecture.
- Error correction controllers (CPUs, GPUs, ASICs, FPGAs) will sit outside the fridge but integrated tightly with the quantum module, with exact architectures still under consideration.
6. Roadmap and Commercialization
- Diraq is targeting a first product release during the first half of 2029: a fully integrated quantum computer module with thousands of physical qubits, enough logical qubits for meaningful problems beyond classical supercomputing.
- Near-term (100–200 qubit) systems will be available in limited cases to select partners and governmental organizations, but the focus is on large-scale, commercially relevant systems.
7. Vision for Quantum Data Centers
- Dzurak envisions thousands of quantum processors integrated into conventional data centers, providing affordable and compact quantum computing alongside AI and HPC for applications such as drug design, materials discovery, and more.
Notable Quotes
"Our technology—the basic concepts go back...over twenty years. But the first demonstrations of spin qubits are really only about ten to fifteen years ago. We modify standard silicon transistors...and then we use the property of the electron known as its spin." — Andrew Dzurak
"We've designed now a system that will go to many millions of qubits that can sit inside one single refrigeration unit, pretty much the size of a rack in a data center." — Andrew Dzurak
"If we want quantum computing to be ubiquitous ... there are going to need to be thousands of quantum computers ... integrated with high-performance computing, GPUs, and so on." — Andrew Dzurak
Episode Takeaways
- Leveraging existing silicon manufacturing and design expertise offers a promising pathway to mass adoption.
- Quantum computing at scale requires not just clever physics, but robust industrial engineering and integration with classical technologies.
- Watch for Diraq’s commercial debut of thousands-of-qubit systems by 2029, poised to play a role in future quantum-enabled data centers.
For further episodes and details, visit www.newquantumera.com or follow on Bluesky @newquantumera.com.