Aerospace giant Rolls-Royce will work with quantum computing software start-up Classiq on a hybrid approach to quantum computing that will speed up testing. This will involve parts of the algorithm run on a quantum and parts on a classical computer. These methods are unlikely to be part of everyday usage for about a decade, the company said, but it is “vital to begin developing algorithms now”.
The long-term partnership will see Rolls-Royce engineers work on algorithms that solve and model fluid dynamics, ensuring they are ready to go when quantum machines reach “advantage” – the stage when they run the same equations faster than the fastest supercomputers.
The algorithms will cover computational fluid dynamics (CFD), which deals with heavy, and complex numerical simulations of fluid and gas. It is “key to improving advanced equipment design and can do this by optimising aerodynamics and thermodynamics”.
One algorithm used in quantum computing that will be adapted to cover fluid dynamics is the HHL, which is designed to solve a linear set of equations very fast. One of its key benefits is that it can be used in a hybrid environment, with developers writing code in Python and sending instructions to both classical and quantum machines.
Deploying the HHL equations and applying them to fluid dynamics will allow for the nonlinear parts of the equations to be solved by a classical supercomputer, and then hand off the linear parts to a QPU which can perform the task significantly faster.
Hybrid approach likely for years
Leigh Lapworth, Rolls-Royce’s quantum computing lead told Tech Monitor this hybrid approach is likely the most common way companies will utilise the benefits of quantum computing, suggesting we will see supercomputer clusters develop featuring quantum processing units (QPUs).
“The first wave will definitely be hybrid classical quantum algorithms,” he says, adding that “this may always be the case with supercomputers developing alongside quantum computers”.
While much of the talk around quantum computing focuses on when quantum supremacy, the moment quantum machines can consistently outperform their classical counterparts, will be achieved, Lapworth believes the reality of quantum for many use cases will be a hybrid approach. He uses the example of very high condition number matrices, “the nasty things that we have to do to model things like jet engines”. He explains: “There won’t be a single moment later in this decade when success is declared [for quantum computers in this area] and that’s it, it will be a hybrid process.”
Content from our partners
How Rolls-Royce will use quantum computing
To get to this point where work can be carried out across quantum and classical machines, an abstraction layer is needed that simplifies the process of writing for quantum computers and can allow one algorithm to run on different hardware. That is the solution Classiq is bringing to the partnership with Rolls-Royce, according to CEO and founder Nir Minerbi.
He says Classiq will deliver and generate optimised quantum circuits that are hardware agnostic, which will allow them to be deployed on different quantum computing platforms that develop in future.
Minerbi told Tech Monitor the goal is to make everything as abstract as possible, explaining that the company hides the underlying manipulation of the atoms in the quantum computer itself, creating a system to manage and automate that process as much as possible so that the top layer can be hardware independent, running in a hybrid environment.
Lapworth says quantum computing will play an important role in helping Rolls-Royce achieve net-zero carbon through ongoing enhancements and sophistication of design simulations.
“The potential of quantum computers to drastically reduce simulation run-times cannot be ignored and the work we’re doing today ensures we will have the capabilities to benefit from quantum advantage when it arrives,” he says.
Being ready for that point takes time. Lapworth believes companies need to invest in the technology today as it can take years to develop appropriate algorithms, and implement new ways of working and systems within a large organisation. “For an industry point of view, it is when, not if,” he says. “When does a quantum computer go faster than a classic computer? Does it allow us to do designs quicker?”
He expects that in five to ten years, quantum machines will be reliable enough to do serious evaluations. “It is a big step to go from evaluations to tools that engineers can use in the business,” he says. “Developing the algorithms and application is where the real value lies today, as we keep applications and algorithms far longer than hardware and to make them rugged for industrial applications take far longer than it takes for the hardware to develop.”
Businesses must develop quantum expertise
Quantum computing adoption is ramping up across industry, with a wide range of businesses starting to experiment with the technology. Tech Monitor has reported that the automotive industry has been an early adopter, with the likes of BMW, Ford and Volkswagen setting up quantum projects.
Classiq’s Minerbi says his team has observed three stages of quantum adoption within businesses so far. “The first one is quantum work,” he says. Building some very, very initial teams that will understand the possibilities, maybe have some hands-on experience with quantum software, with quantum cloud.
“The second phase is what we call quantum readiness, where you have your own use cases, you know how to run them, you are quantum independent. You don’t need some service provider to hold your hand. You have your software stack, your hardware stack via the cloud, and you are pretty much ready.”
Stage three is yet to come, but will be when businesses are ready to scale up the technology, Minerbi says. He adds that early adopters of the technology in enterprise are already five years into their quantum journey, but there is still much work to do before the impact of the technology is widely felt. “The JPMorgan team, the Daimler team, the Exxon Mobile team, these teams are working for five years now and they are just accelerating because there is so much to be done,” he says. “I think that in the next three years, [quantum computing] is not going to change our world. It will bring value to businesses, which will be important.”
He adds: “I think that when we have fault-tolerant quantum computers, which will probably be built by IBM, Google and others, it’s going to be by the end of this decade and at that point things won’t be the same.”