QSimulate Powers High-Throughput Quantum Simulation for Materials Informatics at JSR

QSimulate-MI makes it possible to run high accuracy quantum calculations for thousands of molecules on a daily basis

The QSimulate quantum simulation platform that enables unprecedented high throughput, QSimulate-MI, has been enlisted by JSR Corporation, one of the major players in the semiconductor, display, optical, and polymer materials market, to enable the discovery of novel materials using Materials Informatics (MI) approaches. In this partnership, QSimulate has provided JSR access to its unique automated QM tools on the cloud, making it possible to run high accuracy quantum calculations for thousands of molecules on a daily basis. This, in turn, provides a superior dataset for Materials Informatics, as well as the ability to efficiently supplant that training set as required.

In recent years, the idea that an artificial intelligence (AI) engine for MI, trained using QM-simulated molecular data, could allow rapid prediction of properties from molecular topology has gained traction. If the AI/MI predictions are reliable, an array of relevant material properties can be rapidly assessed, including reactivity, absorption and emission properties, tensile strength, and propensities towards defect and degradation. However, to create a reliable AI/MI engine, a vast amount of high accuracy QM data is required for training, which necessitates a huge number of high-accuracy DFT calculations that have traditionally been both expensive and labor intensive. The QSimulate-MI platform represents a next-generation approach to QM calculations to fully automate the workflow and efficiently utilize elastic scalable computing resources in the cloud with thousands of processors.

Recommended AI News: Sirius Achieves AWS Migration Competency Status

As QSimulate CEO and co-founder Toru Shiozaki notes, “The usefulness of QM for materials design is established, but it is still a slow, reactive process. If we can establish reliable QM-quality predictions using an AI/MI approach on the basis of our high-throughput quantum simulation platform QSimulate-MI, next-generation materials will be developed differently.”

Recommended AI News: Svatantra’s Home Insurance Helping Amphan Affected Rebuild Their Houses

Once an AI/MI model is successfully trained, JSR scientists hope to apply it to such tasks as identifying new materials with desirable properties, replacing old materials with new ones that avoid costly or dangerous reagents, and creating materials better able to hold up under adverse environmental conditions. Yu-ya Ohnishi, Deputy General Manager for JSR, noted, “This is an exciting transitional period in the field of materials design, and collaboration between the quantum experts at QSimulate and the materials experts at JSR holds much promise.” Garnet Chan, Professor at Caltech and QSimulate’s Co-founder agrees. “The stumbling block for QM with respect to maximally impacting materials design has been the computational intensity of these calculations and the complexity of the workflow. QSimulate-MI resolves these issues and represents an important advance in this area.”