Q&A with Dr. Stephanie Simmons
Founder and Chief Quantum Officer at Photonic
Image: BusinessWire
What are the most transformative or exciting applications of quantum computing, generally and in the fields of manufacturing, chemicals and health?
What is exciting about commercializing any branch of physics – which is what we’re doing now with quantum computing – is that the really impactful applications may not be apparent at the start. We likely can’t say right now what the most transformative applications will be, since we’re in the early stages of understanding the opportunities. Think of the early days of transistors. The key application identified for transistors when they were initially developed was expected to be hearing aids. What wasn’t anticipated was the fundamental way in which they would transform electronics, especially modern computers.
With that said, one of the key applications for large-scale distributed quantum processors that we do see is absolutely the materials sciences element because while there is a lot you can do classically, there are some things— from chemistry simulation or chemistry design applications—for which we don’t have the right tools available because classical computers physically can’t solve these problems with a feasible amount of resources. There are some chemical interactions that just require the full complexity of quantum calculations to be able to arrive at the right answer with any degree of accuracy, so it’s going to be exceptional to be able to unlock that capability. Forging down that path and delivering those capabilities will give us a whole new set of tools able to help chemistry and materials science around the work and unlock new opportunities from a design perspective. It’s a new way to approach solutions.
Can you explain the key differences between classical algorithms and quantum algorithms?
To understand the differences in the two types of algorithms, it helps to start with an understanding of superposition, which is a feature of quantum states that allows quantum computers to do things that classical computers can’t do. Superposition is the ability of a quantum system to exist in a combination of multiple states, and the superposed state can be described by something called a wavefunction, which is a function whose value at any given point is related to the probability that the system would be in the state corresponding to that point if measured.
While classical algorithms are designed to use much simpler states, discrete bits of information coded as 0 or 1, quantum algorithms rely on interference of wavefunctions that describe quantum states for efficient calculation, which isn’t a capability available to classical computers. In the course of a quantum computation, wavefunctions interfere in such a way that the amplitudes for incorrect solutions cancel out, while the amplitudes for correct solutions survive. Due to the discrete nature of classical computers, describing an interaction that occurs between wavefunctions with a classical computer is resource-intensive, making it difficult to simulate quantum systems, including quantum computers, effectively.
What are the biggest roadblocks researchers currently face when trying to model materials using classical computing methods?
The biggest challenge right now is that we don’t currently have comprehensive tools that are fit for purpose. With the current tools, small molecules can be modelled, but there is always a tradeoff between the size of the system being modelled and the accuracy of the simulation due to the limitations of classical systems. The utility is even more limited for larger molecules, putting them out of reach for many applications. With catalysts, for example, there is no easy way the quantum mechanics of those chemical processes can be modelled with classical algorithms – it is just outside of the scope of what can be done with classical systems in terms of time and compute power. However, if we can unlock the design tools for ‘all of chemistry’ and ‘all of materials’ – it opens a whole new realm of possibility. In terms of quantum technology that will enable these tools, we need reliable systems with enough capacity to be able to run algorithms that will allow speedups in discovery processes.
We anticipate that even when there are commercially relevant quantum systems, researchers won’t move exclusively to quantum computing. There will be hybrid approaches, where classical and quantum systems are integrated to optimize the portions of the process at which they each excel.
How can quantum computing overcome the limitations of classical methods in material modeling?
Materials modeling for applications including chemistry, batteries, catalysts, drug design, etc., requires simulation of molecules. Molecules are governed by quantum effects, making quantum computers particularly valuable for these use-cases. Quantum computers can achieve a higher accuracy simulation of quantum systems than classical computers can, and with fewer resources. Quantum computers can also tackle the larger molecular systems and reactions that are intractable for classical computing.
There is an entire field dedicated to quantum algorithm design and resource estimation, which focuses on how many qubits and how much time would be required to run an algorithm, how to structure the problem, and what type of qubits/which quantum computing approach might be best suited to that specific problem. Being able to leverage the value of quantum computing for the tasks that classical computers are not able to do, with the compute power and capabilities of classical computers will set up the best scenario to make strides in materials modelling.
How does Photonic contribute to the development of quantum computers generally, and material modeling efforts?
Right now, there are many approaches to quantum computer development using different materials and methods for the building block of the quantum bit, or qubit. Qubits can be considered as equivalent to a bit in a classical computing system. At this point, the industry hasn’t settled on a dominant design, so each of these approaches is contributing insight into important characteristics of quantum computers that help push quantum technology forward.
Photonic is building a quantum computer using the T centre in silicon as our qubit. The T centre provides a unique solution to one of the primary challenges on the road to large-scale quantum utility and adoption, which is scalable entanglement distribution. The T centre is a relatively new qubit candidate for large-scale quantum technologies – in the past five years we have introduced it as a possible qubit, characterized the T centre, and established the properties that can be leveraged for quantum applications. Using this information, we designed the architecture and have successfully demonstrated critical properties including long coherence times, ability to perform local operations, and entanglement distribution.
Our architecture comes from a systems design approach, starting with the objectives of a successful quantum technology – scalability, networkability, integration with existing data centers and telecommunications infrastructure, and working backwards to inform the design.
Entanglement is a quantum resource and is “the glue” of networked quantum computing. Entanglement can be thought of like electricity – it can be distributed and used at different points for a variety of applications. For distributed quantum computing, entanglement is created and then consumed to perform operations by qubits that can be in different physical locations from one another. In our case, we’re prioritizing building a scalable system that can efficiently distribute and consume entanglement using elements from existing telecom infrastructure. Large-scale quantum algorithms running across multiple quantum computers require enormous amounts of distributed entanglement to work well.
As such, our efforts to advance the creation of a horizontally-scalable quantum computer will advance the access of engineers, materials scientists, and others around the world to access the quality and quantity of logical qubits (groups of qubits that work together to increase reliability) required to run the type of algorithms that will advance materials science.
Are there particular quantum algorithms that are showing promise in bridging the gap between the limitations of current technology and the need for new materials?
At the current stage of development, classical algorithms are being developed that can approximate some systems, however known approaches are not able to reach the level of accuracy needed for many of the relevant applications. Trotterization can be used for quantum chemistry applications for small scale molecules beyond the reach of current classical computers, and as we move to larger systems, algorithms like qubitization and variational quantum eigensolvers will become more relevant.
There are many viable algorithms that show promise in materials design, but the biggest challenge the industry is working to overcome is the lack of quantum hardware large enough to run them. At this point, the algorithms are still bound by the capabilities of classical hardware until larger-scale quantum computers become available.
How can researchers and engineers use these algorithms to design and discover novel materials? What does the real-world application look like?
As we don’t have large-scale quantum computers online yet, the answer to the question of ‘What will real world application will look like’ isn’t yet known, but we have some pretty good ideas of how it will start. We know quantum computers are particularly effective at calculating properties of molecules, and we know the problem of designing materials requires searching over a large space of possible molecules, so it is reasonable to expect that a quantum-classical hybrid approach can be used to leverage the strengths of both computing systems – the classical computer can be used to construct candidate materials, and the quantum computer can be used to accelerate vetting candidates to find the best materials from the list constructed by a classical computer.
For those interested in seeing what the development of algorithms for quantum computers will look like, there are quantum development kits available that provide opportunities to write and debug code, choose the type of early quantum hardware you’d like to simulate/run your program on, and do quantum resource estimation. One example is Microsoft’s Azure Quantum Development Kit.
Given the current stage of development, what are the realistic timelines for these applications to become commercially viable?
The timelines for commercially relevant quantum computing vary based on who you ask. We believe there will be some commercially available quantum computing, and access to quantum as a service available in less than five years. Chemistry, and specifically materials science, is one of the applications where there are known use cases, so we anticipate this field will be an early adopter, driving scientific discoveries with global impact.
About Dr. Stephanie Simmons
Dr. Stephanie Simmons is the Founder and Chief Quantum Officer at Photonic, driving the technical vision for next-generation quantum technologies based on photonically-linked silicon spin qubits. She is a world-leading expert in quantum technologies, silicon spin-photon interfaces, condensed matter spin dynamics and control, silicon-integrated photonics, and quantum optics. Stephanie is the co-chair of the advisory board to Canada’s National Quantum Strategy, a Tier 2 Canada Research Chair in Silicon Quantum Technologies, and a Canadian Institute for Advanced Research Fellow in Quantum Information Science. She is also an Associate Professor at the Department of Physics at Simon Fraser University (SFU) where she leads the Silicon Quantum Technology research group.
