The quantum- computer blueprint should be rewritten
Making a difference in quantum computing: Why does the hype come out so fast? Sue Sundstrom, founder and CEO of Riverlane, a pioneer in quantum computers and superconducting circuits
If the field is given time it can be more spectacular than we could imagine at this point in time. The founder and CEO of Riverlane says that short-term hype is a bit high. The long-term hype is not up to par.
Hardware developers can get ParityQC’s intellectual property if they build chips that incorporate the architecture. The company has sold licenses to NEC to build a superconducting quantum chip and has entered several groups to fund the development of quantum technologies after the German government invested 2% of their GDP in quantum technology.
Sue Sundstrom, a start-up coach based in Clevedon, UK, and a judge for The Spinoff Prize 2023, is impressed by what she describes as ParityQC’s analysis of “how previously radically different technologies have been able to get into the market and make money”. She notes a parallel with Arm in Cambridge, UK, a firm which started selling blueprints of chips for reduced-instruction set computers in the 1980s. She likes the hiring of people with commercial expertise. “For quantum companies that is quite rare,” she says.
ParityQC’s architecture helps quantum computers to deal with both limitations. It can do so by changing how the data is stored. Rather than representing values of individual logical qubits as specified by the program, physical qubits record the relationship between pairs of logical qubits in terms of parity. If the qubits in a pair have the same value, then the parity is 1; if the values are different, then the parity is 0 (See ‘Blueprint for quantum computing’).
The spatial properties of qubit are a big issue for quantum computing. The physical processes that link qubits to one another usually occur only over very short distances, such as the overlap of two electron clouds around atomic nuclei or the connection of two superconducting circuits. This means that each qubit typically interacts only with its nearest neighbours, rather than qubits farther away.
But qubits are very fragile. Their states can be disrupted by the slightest amount of heat or other interference. Their durability varies according to the kind of physical qubit used — ions, neutral atoms, superconducting circuits or quantum dots. They might remain intact for a few seconds if they are perfectly isolated or might disappear after milliseconds if they interact with other qubits during a calculation.
The Spin-off Spin-Off Company Lechner, based in the Netherlands, is a spin-off corporation for quantum physicists
The company now has about 30 people working for it. It landed a contract worth tens of millions of euros from high-tech manufacturers. According to Lechner’s co-chief-executive Magdalena Hauser, this early success — combined with grants from the European Union and the governments of Austria and Germany — has meant that the company has avoided having to drum up support from venture capitalists. “We made revenue from the start,” says Hauser.
Just six months after he filed a patent, he received an offer from a large technology company for his intellectual property. (Lechner declined to reveal the company or the size of the offer.) It was suggested that it might be better to try to get the rewards directly in order to maximize commercial potential. So he and his colleagues at the University of Innsbruck decided to reject the offer and set up a spin-off company. ParityQC is a finalist in The Spinoff Prize.
The physicist at the University of Innsbruck in Austria, Lechner, discussed the idea with his colleague, but they didn’t think it would work. Over the next two years he kept turning the idea over in his mind, and says it became an obsession. After 3 a.m. in a hotel room, he had a flash of inspiration that might mean that his approach will work after all.
Many think that the first commercial applications of quantum computers will be speeding up and gaining better control of reactions. “If anything is going to give something useful in the next five years, it will be chemistry calculations,” says Ronald de Wolf, senior researcher at CWI, a research institute for mathematics and computer science in Amsterdam. Shintaro said that the low resource requirements made it possible. “This would be possible using quantum computers with a relatively small number of qubits,” he says.
The problem is compounded by the difficulty of building the hardware itself. Quantum computers store data in quantum binary digits called quantum bits, or qubits, that can be made using various technologies, including superconducting rings; optical traps; and photons of light. Some technologies require cooling to near absolute zero, others operate at room temperature. Hensinger’s blueprint is for a machine the size of a football pitch, but others could end up installed in cars. There is no agreement about how to measure the performance of quantum computers.
Whatever the design, the clever stuff happens when qubits are carefully coaxed into ‘superposition’ states of indefinite character — essentially a mix of digital ones and zeroes, rather than definitely being one or the other. The evolution of these superposition states can be done with a quantum computer. The quantum rules of this evolution allow the qubits to interact to perform computations that are, in practical terms, impossible using classical computers.
For instance, Nicole Holzmann, who leads the applications and algorithms team at Riverlane, and her colleagues have shown that quantum algorithms to calculate the ground-state energies of around 50 orbital electrons can be made radically more efficient5. The previous estimates of the runtime were more than one thousand years old. But Holzmann and her colleagues found that tweaks to the routines — altering how the algorithmic tasks are distributed around the various quantum logic gates, for example — cut the theoretical runtime to just a few days. That equates to a speed gain of five orders of magnitude. “Different options give you different results,” Holzmann says, “and we haven’t thought about many of these options yet.”
At IBM, Garcia is going to exploit the gains. The potential quantum advantage isn’t limited to calculations involving a vast array of molecules in many ways.
One example of small-scale but classically intractable computation that might be possible on a quantum machine is finding the energies of ground and excited states of small photo active molecules which could improve manufacturing methods for chip and drug design. The singlet and triplet states of a single oxygen molecule are of interest to battery researchers.
Six quantum simulations of the H3S+) were published in February. The molecule is related to a photo-acid generator used in the production of light-generated acid. Understanding its properties could make the manufacturing technique more efficient. The H3S+ ran on IBM’s falcon processor, which has just 27 qubits, but it looked impossible when the team began the work.
Part of the IBM team’s gains are the result of measures that reduce errors in the quantum computers. In some cases, such as error mitigation, noise is cancelled out using a similar method to those in noise-cancelling headphones, and in other cases, like entanglement forging, it’s possible to separate out part of a quantum circuit without losing quantum information. The technique which double the quantum resources was actually invented last year.
In exploratory work with South Korean car manufacturer Hyundai, Jungsang Kim at Duke University in Durham, North Carolina, and researchers at the firm IonQ in College Park, Maryland, developed quantum machine-learning algorithms that can tell the difference between ten road signs in laboratory tests (see go.nature.com/42tt7nr). Their quantum-based model used just 60 parameters to achieve the same accuracy as a classical neural network using 59,000 parameters. Kim says that we need far fewer training iterations. “A model with 59,000 parameters requires at least 100,000 training data sets to train it. Your training becomes very efficient because of the small number of parameters with quantum.
That’s not a problem, Simmons adds: Silicon Quantum Computing has patient investors. So, too, does Riverlane, says Brierley. “People do understand that this is a long-term play.”
“I’m not trying to reduce the amount of work done, but we are surprised by how much we’ve done.” said the senior research manager for quantum applications and software at IBM.
If you believe the hype, computers that exploit the strange behaviours of the atomic realm could accelerate drug discovery, crack encryption, speed up decision-making in financial transactions, improve machine learning, develop revolutionary materials and even address climate change. The surprise is that those claims are now starting to seem a lot more plausible — and perhaps even too conservative.
