Effective 10/13/2023, we’re pleased to share that the Microsoft Quantum Inspired Optimization (QIO) solver is now available on GitHub under the MIT license in collaboration with KPMG. We’re excited for vibrant, community based QIO innovation to continue and generate progress and applications. Access here on GitHub.
Whether you’ve noticed or not, you probably spend at least some part of your day staring into an OLED (organic LED) display, as they are found in smartphones, tablets, televisions, and computer monitors, to name just a few applications. OLED displays use organic carbon-based molecules to generate light of different colors under an applied electrical current.
Breakthroughs in displays, and most other technological fields, can be traced back to advances in materials science that enable the discovery of advanced materials with unique properties. However, designing new materials with specific desired attributes is extremely difficult because small changes in the structure of atoms that make up a material can dramatically influence its properties.
Computational chemistry simulations can help accelerate the design of new materials, by providing a better understanding of these structure-property relationships. These simulations pose a huge computational challenge because of the complexity of simulating the characteristics of quantum physics, which governs the interactions between atoms, but we now have the compute power to solve some problems that previously seemed intractable on classical hardware, leading to breakthroughs in new materials discovery.
OTI Lumionics has developed a fast materials design approach, tailored to OLEDs and other electronic materials, that consists of a combination of machine learning techniques, computational chemistry simulations, optimization, rapid synthesis, and closed-loop feedback from testing of new materials in pilot production. They work with the largest electronics companies in the world to design new materials that are mass-production ready, enabling the next generation of exciting consumer electronics.
One application of OTI Lumionics materials, that has been designed using this approach, is in transparent displays, which will soon be available in smartphones, helping to hide the array of sensors and front-facing camera under the display. When you are “heads-up driving” – viewing your speedometer and mileage on the windshield of your car – you are looking at another application of this technology.
Instead of using a traditional approach to materials discovery which requires synthesizing and testing thousands of variations to find the suitable candidate, OTI Lumionics has developed software tools to simulate and predict the properties of new materials, allowing a larger pool of candidates to be screened than could otherwise be synthesized and tested. Thus, new materials that meet the precise requirements of the largest electronics manufacturers can be “designed” rather than discovered by chance.
The slowest and most expensive part of the workflow is the computational pipeline – the bottleneck on available hardware when running extremely large simulations, which scale exponentially with size. In addition, some simulations are so compute-intensive that they are literally unsolvable with today’s classical computers. The trade-off between simulation accuracy and compute-intensity is thus a major bottleneck in using a computational approach for commercial size problems.
To get around this bottleneck OTI Lumionics has been investigating quantum computing as a potential candidate to help accelerate computational chemistry simulations of new materials. Since many structure-property relationships of materials are governed by quantum physics, quantum computing, which uses quantum mechanical effects to perform computations, is a natural candidate to simulate these systems more accurately.
“Quantum computing has the potential to revolutionize materials design, by enabling highly accurate simulations that could otherwise not be solved on classical hardware. Unfortunately, current gate-based quantum computing is far from being powerful enough to simulate commercial-sized problem,” said OTI Lumionics Head of Materials Discovery, Scott Genin.
Using Azure Quantum and quantum optimization solutions running on classical hardware, Quantum Inspired Optimization (QIO) can enable quantum methods for materials simulations that yield more accurate results.
Scott Genin again: “In the field of computational chemistry, high accuracy property prediction is considered to be very difficult; in fact, some computations are nearly impossible on today’s classical hardware. We have developed new methods, that allow quantum computing algorithms for computational chemistry simulations to be represented as binary optimization problems. Running our quantum computing methods with Azure Quantum optimization solutions, we are getting results that are more accurate than other algorithms.”
As an early adopter of quantum computing, OTI Lumionics has invested in a team of quantum chemists, computer scientists, and software engineers to develop their own quantum computing algorithms and software for materials design, and have made significant theoretical and practical advances in the field. With their algorithms now running on Azure Quantum, OTI Lumionics is able to demonstrate meaningful results on commercially relevant sized problems, today. For example, by using Azure Quantum’s optimization tools in their pipeline, OTI Lumionics successfully performed a complete active space configuration interaction simulation of an archetype green light emitting OLED material – Alq3 [Tris (8-hydroxyquinolinato) aluminum].
“We have designed our solver platform in Azure Quantum with customers in mind,” said Microsoft Principal Research Manager, Helmut Katzgraber. “Our quantum solutions on classical hardware do not have the limitations of other solvers and optimization hardware and are driven by some of the most powerful algorithms currently available, while being easy to use as there is no need to tune parameters. “
To give you an idea of the computational savings the same simulation of Alq3 would require 42 error-corrected qubits on gate-based quantum hardware. Mapping the problem to an industry-standard quadratic unconstrained binary optimization (QUBO) using OTI Lumionics reparametrization would require a quantum annealer (or QUBO solver) that could handle 58,265 variables. Solving a QUBO problem with this many variables is intractable, and even an equivalent simulation of Alq3 using standard classical computational chemistry software would require a supercomputer. In contrast, using Azure Quantum, the higher-order binary problem can be handled natively, meaning that this problem only requires 132 variables on classical hardware to perform the simulation.
“The fact is that we have compelling results that show that we can start using quantum solutions for commercial problems in a matter of months, not years,” said OTI Lumionics Co-founder and CEO, Michael Helander. “Using Azure Quantum, we now have the potential to dramatically increase the accuracy and throughput of the computational chemistry simulations that underpin our entire materials design workflow.”
Using Azure Quantum, OTI Lumionics can open their computational pipeline to run more accurate simulations at significantly higher speeds, which could ultimately lead to timelier and lower cost materials design, and thus better OLED displays.
We are excited to be working with OTI Lumionics in helping them find breakthrough discoveries in materials through quantum computing and Azure Quantum.