The simplest way to access quantum computational power at a distance is through cloud technologies. This is already a service provided by several companies, with time being available on a scale of minutes to hours, while prices range from free to several thousands of dollars per hour. Given the high costs of quantum computers, the cloud delivery strategy (1) expands the market for quantum computers and (2) gets potential buyers of quantum computers used to using quantum computers.
We are already seeing major firms in the quantum computing space take the cloud delivery strategy, although with different emphasis on (1) and (2). The Exhibit below provides an illustration of where we are with “quantum clouds.”
Selected Quantum Cloud Services: Mid-2019
|IBM Q||Hub at Keio University launched in May 2018, which is servicing companies like JSR, Mitsubishi UFJ Financial Group, Mizuho Financial Group, and Mitsubishi Chemical. IBM still gives away time to researchers, but charges corporate customers|
|D-Wave||NASA has signed a contract with Ford for $100,000 for access to NASA’s D-Wave 2000Q device located in Silicon Valley. Meanwhile, D-Wave’s LEAP program gives developers access to its systems for developers. Leap, for example, offers 1 free minute per month, and one hour costs $2000|
|Rigetti||Rigetti has Quantum Cloud Services by Rigetti.|
|Quantum Simulation Clouds||Access to online quantum simulators also exists. This is already provided at universities like Bristol and Tsinghua, and also by QuTech and Google.|
Source: Inside Quantum Technology
For the most part, these services simply provide access to quantum resources via the regular Internet. As such this kind of quantum service is really just one small chapter in the long history of sharing expensive computers over some kind of network. But we think that there are important limitations to this kind of cloud access.
Many end users – actual and potential – for quantum computers need enhanced privacy and in the kind of quantum cloud that currently exists, the server can see all the client’s operations being performed, and the client can’t be sure that the algorithms weren’t tampered with. The military, government and industrial customers that are turning quantum computing into a business, not just a research program.
Attempts are being made to develop more secure protocols where the client is purely classical, but we think that another opportunity is to develop a secure approach requiring minimal quantum processing capabilities, like preparing and measuring single qubits. With QKD already being viable a credible technology, we don’t believe this to be a major roadblock. And this opportunity could provide QKD providers with a new revenue stream.
QKD providers are the obvious choice since their devices already incorporate single qubit preparations, but there might be sufficient differences that there will be places for new players to emerge on the market. All in all, we believe the real value of quantum computers will shine once their capabilities are made available to the world in an effective and private way, which requires hardware advances on both the server and the client sides.
Overall, the opportunities in this space therefore begin to look like what is shown in the Exhibit below.
Selected Quantum Cloud Privacy Technologies
|Privacy Technology||Market Potential|
|Classical Cloud||What is available now, but quite insecure. Market disadvantage in a market where prospective customers value security.|
|New Algorithms||Essentially, the cloud version of PQC. Has some potential, especially in the short-to-medium term.|
|Hardware-Based Solution||Some kind of versioned QKD-like approach. Demanding customers will probably want this in the long run.|
Source: Inside Quantum Technology
If you’d like to know more about quantum computers and the future of quantum clouds, visit our conference in the Netherlands in October 2019.