A Cryogenic IC May Be the Answer to Quantum Computer ScalabilityFebruary 04, 2021 by Jake Hertz
Temperature has been a long-standing issue in scaling quantum computing. Now, researchers in partnership with Microsoft may have discovered the key to scalability: a cryo-CMOS chip.
One of the many challenges facing the future of quantum computing is scalability. While researchers predict that quantum computers will need to operate on hundreds of thousands of qubits, current quantum computers still work on double-digit qubits.
Current state and trends in the number of qubits achieved. Image used courtesy of Edoardo Charbon
Researchers at the University of Sydney, in collaboration with Microsoft have now developed a technique that they believe will allow for the needed scaling of quantum computers.
Challenges to Quantum Scalability
Quantum scalability is hindered largely by the two conflicting constraints.
The first constraint is that to manipulate qubits with control signals, contemporary architectures require each qubit to be controlled individually by external circuitry. These control signals are produced by classical—not quantum—electronics.
The second constraint is the fact that quantum computers must be kept at cryogenic temperatures (near absolute zero) to behave deterministically in the world of quantum mechanics.
Typical quantum computer block diagram with temperatures. Note that 4K is room temperature. Image used courtesy of Edoardo Charbon
This is where the fundamental incompatibility occurs: classical electronics must be used to individually control the qubits, but the qubits must operate at near absolute zero. Classical electronics at cryogenic environments are still in their infancy. Instead, designers of quantum computers operate the classical electronics at room temperature and use long, clunky cables to interface the two.
This is a large factor limiting quantum computer scalability. In fact, a recent research experiment required around 200 wideband coaxial cables, 45 microwave circulators, and racks of room-temperature electronics to control just 53 qubits. How can researchers take the next step to control thousands of qubits?
A Cryogenic CMOS IC Controls Thousands of Qubits
For over four years, researchers at the University of Sydney have been working with Microsoft on a solution to this problem and have announced promising developments this week.
Their solution is to eliminate the distance and temperature difference between the classical-quantum electronics. They’ve done this by developing a CMOS-based qubit-controlling IC designed to operate at cryogenic temperatures, down to 100 mK.
The single IC is capable of controlling thousands of qubits while only requiring two wires as input, eliminating the massive infrastructure needed by conventional quantum computers.
The cryogenic IC in a dilution refrigerator. Image used courtesy of the University of Sydney
Along with increasing scalability, this solution—bringing classical electronics closer to quantum electronics—has the added benefit of reducing power consumption and upping speeds.
The Cryo-CMOS Chip Combines Analog and Digital
As explained in the researchers' published article in Nature, the cryo-CMOS chip consists of both digital and analog blocks working in unison. The digital circuit blocks are utilized for communication, memory (in the form of a 128-bit register), and autonomous operation of the chip via a finite state machine consisting of roughly 100,000 transistors. This section also contains a configurable ring oscillator.
The IC’s floor plan. Image used courtesy of Pauka et al.
On the analog side is the charge-lock fast-gate (CLFG) cells, which work using switched capacitor technology to generate dynamic and static voltages necessary for controlling the qubits. The CLFG blocks leverage the low leakage of devices at cryogenic temperatures, utilizing switched-cap-based techniques to generate and move around charges as needed.
Working with Microsoft, the Sydney researchers anticipate that the chip won't remain just in academia, but will be fully manufactured. The IC was built on 28nm fully-depleted silicon-on-insulator technology.
Closer to a Quantum Future
Speaking to the future of the project, head researcher David Reilly says, "We are just getting started on this new wave of quantum innovation." He continues, "The great thing about the partnership is we don’t just publish a paper and move on. We can now continue with the blueprint to realize quantum technology at the industrial scale."