What can we learn about the Future of Quantum Technology from the past? Here we look at the recent computing past and draw some parallels which could provide the future trends of the quantum future and help everyone navigate the excitement in the quantum space.
Before there were desktops, there were mainframes. These giant room-sized machines powered calculations such as the NASA moon mission. But they were also used to perform more mundane work, such as billing you a monthly electricity statement. Every quantum computer you use requires special hardware, and many require refrigeration for superconducting Qubits. In the 1960s, the IBM 360 was one of the most popular mainframe machines. Today IBM, through its Quantum series of machines, is allowing users from around the globe to run quantum circuits on its hardware akin to the 1960s mainframe equivalent, taking the space of a room.
IBM 360: One of the most popular mainframes from the 1960s and 1970s
The silicon revolution kicked off with the integrated circuit. The ability to assemble transistors into larger devices is the root of the modern processor. Quantum processors are also being developed from an assembly of Qubits. Just as their classical analogue permitted, quantum processors will likely prove a useful concept. Companies and researchers are looking to scale the number of Qubits from tens and hundreds to thousands and millions. Expect the equivalent of the IC or Integrated Circuit for Quantum Computing.
Room Sized Machinery from the 1970s parallels with Room Sized Quantum Computers of Today
There is some radical tech. Before the transistor, there were valves, and each of these could be thought of as the predecessor to the transistor; just as there were many types of valves and devices that had very different operating behaviours and designs, so too due to all the competing qubit technologies today. Before the industry accepted silicon and semiconducting devices, there were many valve alternatives, such as decatron and more.
There are many qubit technologies available today, and it’s yet to be decided what or which quantum technology will win out.
Computing technologies are dominated by just a few companies that create the hardware and software that dominate our computing landscape. Few companies can afford the costly R&D costs, ultimately precluding many from the marketplace. Whilst it’s early days for the quantum industry, producing qubit devices is far from a bedroom pursuit. We’ll likely see the same pattern of innovation and consolidation in the Quantum space, where larger companies buy into creation from smaller startups. For example, Cambridge Quantum Computing joined forces with the Quantum division of Honeywell (the conglomerate) to found Quantinuum, and whilst there are no further mega-mergers quite yet, we do expect to see larger, more established players buy out smaller businesses to boost their offering.
The operating system or OS meant that computer programmers didn’t need to know every different machine intimately. Accessing a disc could be achieved with a universal command. On top of the OS is a high-level language such as C or C++ that allows even more abstraction from the hardware. We are already seeing this trend with the evolution of two key innovations; the Quantum OS and portable Frameworks equivalent to the higher-level language of the past. No more do programmers have to work in machine-specific languages such as Assembly and care about memory addressing and machine features. Frameworks like Qiskit allow precisely the same abstractions. Still, if you want that lower-level language fix (Brian Siegelwax), you can program in QASM, which is Quantum Assembly. Riverlane is a British company doing exciting things, and it is building a Quantum Operating system, which should be technology agnostic and allow developers/programmers to build more efficiently. Deltaflow.OS intends to address several challenges in quantum computing, such as: Improving interoperability between different quantum computing architectures, Optimizing performance and Providing a foundation for a quantum software industry. The latter is crucial since this parallels with the past as standard operating systems kickstarted a revolution in computing with DOS, Windows, Linux and MacOS.
A brief look at the Future, maybe…
Predictions are notoriously unreliable, but we have highlighted that we think there are some key parallels between the large-room-sized conventional computers, which shrunk to desktops and then into our pockets and even wrists. Whilst quantum computers are room-sized, it’s not even necessary that they shrunk in stature, as cloud-based systems enable developers to program them from afar.
The development of silicon microprocessors followed the integrated circuit invented by Jack Kilby and which went on to create the computer revolution. We don’t think quantum processors are pretty at that stage, but they are close. Are we in the quantum equivalent of the 1950s?
A quantum operating system is likely to be crucial. We already have great development tools and languages such as qiskit, cirq and Q#. But as hardware improves and qubit count improves, expect decoupling between hardware and software. Amazon AWS provides some of that glue to be hardware agnostic with its Braket service, allowing users to select and switch between hardware makers. But they’ll be a space for OS’s which might help accelerate the players in the hardware space.
Looking back, we think you can see the future and unpick those trends that will dominate the future.