An EU-funded research team raises big hope that quantum computers can finally be housed in normal-sized server racks outside controlled environments
As we are mostly aware by now, quantum appears to be the future of computing. Qubits are set to achieve computational standards that are way beyond the capabilities of conventional bits and bytes. Both the currently operational quantum computers – be it the modest-sounding 53 qubit Google product or the 65 qubit IBM innovation –can perform within a few minutes computations that would take ten thousand years for a conventional processor to complete. And the industry is abuzz with news of the one-million-qubit quantum beast that IBM is developing right now.
But word is also spreading how the power and magnitude of a one-million-qubit quantum computer would require super cooling infrastructure and gigantic housing facilities. Press releases have already mentioned “Goldeneye” – the 10 feet tall and 6 feet wide massive refrigerator that IBM is manufacturing to keep their million-qubit wonder at around 15 millikelvin (-459 degrees Fahrenheit) – something even colder than outer space!
Amidst all this noise about magnitude and temperature, a recent development has provided a new direction to building practical quantum computers. A team of researchers at the University of Innsbruck in Austria have taken one big leap towards bringing quantum computers out of the lab. Under the aegis of an EU-funded project called AQTION, the scientists have designed a prototype quantum computer that is compact enough to fit in ordinary server racks. The complete device consists of a fully functional ion trap quantum computer that only requires a single wall-mounted power plug and can be snugly housed within two 19-inch racks, typically used to place servers in data centres around the world.
The development of this prototype signals a freeing of shackles in a scenario where quantum computers were thought to be mainframe-equivalent concept-machines that could be installed only if customised infrastructure is available. Quantum computing was increasingly being regarded as a set-up involving nearly 50square meters of laboratory space. Acompact housing with a single-point power outlet means quantum computers could be visualised as a more accessible solution suitable for widespread use. And that was exactly what the researchers had set out for. The EU-driven AQTION is a €10 million project with a brief to design a compact ion-trap quantum computer which fulfils all requisite industry standards and can run in normal room environment just like any conventional computer.
The scientists had their task clearly cut out. Fitting a quantum computer in a pair of 19-inch server racks meant every individual component of the machine had to be physically scaled down. And that also included the ion trap processor and the vacuum chamber. There lied the biggest catch – because reducing them in size considerably could mean substantial impact on computing performance. However, the team announces that hurdle has been overcome and the prototype was performing without compromise. Measurements revealed that the system’s performance and error rate were on par with lab-based implementations. The prototype also remained stable outside lab-environments and operated without interruption from external disturbances. The research team could individually control and entangle up to 24 ions. They estimate that this would go up to 50 individually controllable qubits by 2022.
The AQTION quantum computer unit snugly housed in two 19-inch server racks.
Image courtesy: University of Innsbruck
The prototype is still work-in-progress. The hardware and software would be further upgraded before it is released online. Researchers will access the device over the Cloud to test quantum algorithms on a hardware-agnostic quantum computing language. So you might not be getting your own desktop quantum machine anytime soon. But now there is hope that it is possible for quantum systems to effectively step out of the Goldeneye refrigerator. AQTION has indeed proved that compactness does not have to come at the expense of functionality.