Sample of a Few Materials from Phasecraft’s Materials Quantum Modeling Complexity Database

One of the most promising areas for near-term commercial application of quantum computing is in the area of materials modeling for discovering new materials and drugs. Applications in this area may only require a few thousand qubits as opposed to the millions of qubits needed for other applications. Progress in reaching this goal is coming from two fronts. The first is the continued improvement in hardware capabilities as we report frequently in these pages. But the other is progress in creating better algorithms that can perform the needed computation with a fewer number of qubits or gate depth.

Now, Phasecraft has published in Nature Communications a paper that describes their approach which reduces the gate depth by roughly 5 or 6 orders of magnitude for a Trotter layer of time-dynamics simulation. Their software can support either a  variational quantum eigensolver (VQE) to approximate the ground state of a Hamiltonian variational ansatz19 or a time dynamics simulation (TDS) algorithm. As an example, the company’ assets that their algorithm can simulate a lithium copper oxide (Li2CuO2) material with 410,000 gates versus a baseline of about 1.5 trillion with a more traditional quantum algorithm. With reductions like these, one of their quantum material simulation algorithms could be expected to run successfully for many materials on quantum processors that we expect to be available within the next few years.

Phasecraft has used its algorithms to create a Materials Modeling Complexity Database that compares the number of qubits and gate levels needed to simulated 46 different materials between a more traditional baseline approach and their approach. The materials selected cover a wide range of interesting application areas including batteries, fuel cells, sensors, transistors, photovoltaics, and many more. The number of qubits needed is roughly similar between the two approaches the number of gate levels needed show the 5-6 orders of magnitude improvement. The qubits shown in the table could either be logical qubits on an quantum computer with error correction or possibly physical qubits with an error rate of under 10-4 or 10-5 with corresponding coherence times.

This development was part by grants from the Innovate UK program and the National Quantum Computing Centre (NQCC) grants. The company also collaborated with the Scientific Computing Department at the  Science and Technology Facilities Council (STFC) for this research. Phasecraft currently has about 25 employees and closed a £13 Million ($16.5M USD) Series A funding last year. They work with quantum hardware providers Google AI, IBM, and Rigetti and collaborate with a number of end users including BT, Oxford PV, Johnson Matthey, and Roche.

Additional information about this development can be found in a LinkedIn post here. The Nature Communications paper can be found here. The Materials Modeling Quantum Complexity Database is available here and a description of the methodology used to create the database can be seen here.

January 25, 2024