![](https://www.techmonitor.ai/wp-content/uploads/sites/29/2023/02/shutterstock_2136788203.jpg)
Quantum computers are fragile, temperamental machines – even the handfuls of qubits inside the most advanced versions, after all, are liable to be toppled on enlightenment’s shore by tidal waves of noise. The last thing anyone should want, one would therefore think, is to make these machines accessible through the cloud. Available to any company willing to pay to harness their raw computational power, it would be reasonable to get a little nervous at any mistakes propagating through corporate calculations and design specifications, slackening complex supply chains or undermining the structural integrity of jet engines.
![Cloud computing is making it easier for developers to take advantage of the power of quantum hardware to solve complex problems (Photo: Gorodenkoff/Shutterstock)](https://techmonitor.ai/wp-content/uploads/sites/4/2023/02/shutterstock_2136788203.jpg)
That might have been true a decade ago, when error correction measures were in their infancy and the number of rational qubits supportable inside a quantum computer could be counted in only single digits. Recent years, however, have seen remarkable progress on both fronts, affording the quantum computing field its first opportunities to craft practical applications for the medium in a range of sectors. In that sense, making these machines available through the cloud is a boon. A quantum cloud would, advocates argue, prevent the need for interested companies to invest huge sums to acquire and maintain their own quantum computers, thereby speeding up research into and the deployment of unique, quantum-powered solutions to their own, niche set of problems.
In that respect, says Richard Hopkins, an IBM Distinguished Engineer and fellow at the UK’s Royal Academy of Engineering, the cloud “has been absolutely vital,” resulting in “thousands of thousands of papers from academics and researchers and kickstart[ing] a whole ecosystem around quantum computing.”
It was, in large part, a revolution sparked by IBM’s decision to make its quantum computing hardware accessible through the cloud as early as 2016, before launching its Qiskit open-source software deployment kit the following year. Since then, most of the hyperscalers have joined the race to offer an efficient, quantum cloud solution, with dozens of startups and scale-outs building their own solutions on top, like so many remora fish tentatively following their chosen shark. There are also benefits for many other stakeholders further downstream.
“Making quantum computers easily available to anyone via the cloud demonstrates that quantum is real because now anyone can run a quantum program with a few minutes and a credit card,” pronounced IonQ CEO Peter Chapman, in an interview exploring his own firm’s collaboration with Google Cloud. “This democratisation of access is core to realising the promise of quantum.”
Making it rain in the quantum cloud
The applications that have arisen from access to quantum clouds vary widely in scope and ambition. Companies in material sciences, manufacturing, healthcare and AI, for example, are making active use of quantum computing through cloud platforms, including Bosch, which recently developed new energy storage and fictional materials through quantum simulations run on IBM machines. Volkswagen is similarly using the Google Quantum AI cloud to improve traffic flow optimisation, while Biogen has harnessed Microsoft’s Azure quantum to accelerate drug discovery and development.
Few of these achievements could have been possible without quantum cloud facilities, argues Piers Clinton-Tarestad, a partner and quantum computing leader at EY, not least given the expense and expertise required to run one’s own quantum computing initiative. Quantum clouds, by contrast, simplify matters enormously in terms of cost but also by providing a new level of abstraction to the process. “That makes it easy for people to experiment through known interfaces,” says Clinton-Tarestad, “and to integrate quantum computing into solutions such as Python language models.”
Providing open and easy access to quantum computing power has led to a flurry of new algorithms, some commercial and restricted to a single company, others open-source. It also has the potential to become a lucrative side hustle for your average tech giant. IBM, for example, claims more than 400,000 active users for its quantum cloud.
“The next step is that we're going to have essentially serverless access to these things in real time,” said Hopkins, with quantum computers already in the same halls as GPUs and more traditional hardware. The end goal, he adds, is for customers to have an almost unthinking attitude toward solving complicated problems in the quantum cloud, and for visions of the hardware “to disappear below the waves.”
Hardware dysfunction
It isn’t just IBM seeing commercial use cases for its quantum cloud platform. Microsoft has been working with clients developing algorithms on its own Azure Quantum cloud. "We've seen exciting work already from customers and partners in traffic optimisation, financial modelling, transportation and logistics, materials design, and more," said Julie Love, senior director at Microsoft Quantum speaking to ZDNet. "I'm most excited to see what new ideas developers come up with once they've had the tools and solutions in their hands, particularly for solutions to our biggest challenges in climate and the environment."
Challenges nonetheless persist in expanding the reach of quantum clouds beyond niche industrial use cases toward wider applications. It’s a story, as ever, of hardware catching up to ambition, a concern that IBM, Google and others are attempting to address by scaling up their own quantum processors. Error handling, too, needs to keep pace.
“What we’re doing now is we’re introducing layers of error mitigation in quantum computers and improving the quality of our qubits” says Hopkins about IBM’s work in this area. Eventually, he explains, “we will need logical qubits that are made of thousands of qubits… but in the meantime, through these mechanisms, we can reduce the noise, mitigate for it.”
The idea is to create a system that, while not necessarily capable of running at quantum supremacy levels, can run sufficiently accurately and at a speed that makes it more cost-effective than a classical supercomputer. “That’s the goal,” says Hopkins. “To make this stuff financially attractive for our customers to get market advantage.”