Quantum chemists will soon be able to test their algorithms on a trapped-ion quantum computer that automatically suppresses its own errors, without needing to hire a specialist team to build custom error-correction tools. Classical computers cannot accurately simulate large molecules because they must make too many approximations. Quantum computers could do this precisely, but today’s machines are too noisy to run useful chemistry calculations. Error-mitigation methods exist, but they consume large amounts of quantum processing time and must be tailored to each specific algorithm and hardware platform. Researchers currently lack a ready-made tool that lets them focus on designing better chemistry algorithms rather than on building error-mitigation software from scratch. The Q-CHEMION consortium will build a complete software-hardware stack that combines Oxford Ionics’ high-fidelity trapped-ion qubits with Qedma’s error-suppression middleware, optimised specifically for quantum chemistry. If successful, the platform will let researchers simulate molecular processes—such as the first steps of vision in retinal proteins—at scales far beyond classical supercomputers. This could accelerate the discovery of new drugs and energy materials by providing accurate theoretical models where none currently exist.
View original technical description
The theoretical modeling and computational simulations of molecular systems is essential to understanding the mechanisms behind a number of chemical and/or biological processes, for example, the way in which our eyes react to light. Classical simulations of such quantum systems require many approximations of the system behavior and therefore have limited accuracy. The development of quantum chemistry algorithms for quantum computers to describe molecular systems can overcome hampering bottlenecks in system size, simulation time, and accuracy. This will pave the way for much-needed theoretical support to the development of, e.g., new drugs and energy materials. However, development of quantum algorithms for quantum chemistry problems is facing a great challenge. At large scales, the performance of these quantum algorithms cannot be investigated using simulations on classical computers. Additionally, execution of quantum algorithms on quantum hardware suffers from the acute problem of noise and errors on these devices. Error mitigation (EM) methods are geared to significantly reduce errors without requiring any overhead in the qubit count, yet they require significant QPU runtime. Thus, it is essential that they are tailored to the specific hardware and use-cases. Researchers developing quantum chemistry algorithms do not have dedicated tools for executing their algorithms on low-noise quantum hardware with EM. This slows down the progress of research and requires the research teams to hire talents to develop in-house methods, which do not require a quantum computer to bridge this gap instead of focusing their resources on the development of quantum chemistry algorithms. The consortium hereby proposes Q-CHEMION, a full-stack quantum solution that will allow researchers to test and benchmark quantum algorithms, eventually going far beyond the scales which can be simulated by classical supercomputers. The stack will include: 1. Use cases: Prototypes of real-world use cases within life and materials science, for example, processes in light-sensitive retinal proteins, like rhodopsin, which constitutes the first few steps in vision. 2. EM middleware: Qedma's Quantum Error Suppression and Error Mitigation (QESEM) product will be tailored and optimized for quantum chemistry algorithms on OI's cutting-edge hardware. 3. Hardware: Oxford Ionics' trapped ion-based quantum systems are driven by their unique electronic qubit control system. They have achieved record-setting qubit control fidelities while maintaining the scalability of their architecture systems. Algorithms presented by DTU, UCPH, and MQS will be run on OI hardware using algorithm-agnostic Q-CHEMION.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know