The Materials Project is the most-cited resource for materials data and analysis tools in materials science. It has been cited over 32,000 times in peer-reviewed studies, driving advances in fields like batteries and quantum computing. Serving more than 650,000 registered users, the project democratizes materials knowledge and fosters collaboration.
Materials Project: Impact on Materials Science Research
The Materials Project accelerates materials research by utilizing high-throughput computational modeling at NERSC to screen vast libraries of materials. This process calculates properties using advanced methods, validated by experiments, allowing for rapid evaluation of hundreds of thousands of compounds. Researchers can then leverage this data to design materials more quickly and build machine-learning models predicting materials behavior for specific applications. This platform provides standardized, curated datasets – including electron density information – ideal for training AI systems and benchmarking performance. The extensive preparation of data eliminates months of cleaning and assembly, letting scientists focus on algorithm development and discoveries.
During recent restrictions, the Materials Project’s digital tools enabled continued research, now supported by a cloud infrastructure guaranteeing 99.98% uptime and serving a rapidly growing user base.
Kristin Persson’s Vision for Accelerated Materials Design
Kristin Persson’s vision centers on accelerating materials design through a freely accessible, automated screening tool. The Materials Project, founded in 2011, aims to empower both academic and industrial researchers to rapidly develop materials—particularly for energy technologies—by offering the largest collection of materials properties. This platform utilizes high-throughput computational modeling at NERSC, calculating properties and validating them with experiments to speed up the discovery process. Persson and her team intentionally built the Materials Project to be “AI ready”, anticipating the need for large, well-curated datasets.
The platform now delivers approximately 465 terabytes of data to users, facilitating machine-learning applications without extensive data preparation. This focus on machine learning has allowed researchers to bypass repetitive testing and focus on innovation, ultimately unlocking advancements in fields like batteries and catalysts.
650,000 Users & 32,000 Citations Demonstrate Database Growth
With over 650,000 registered users, the Materials Project sees approximately 5,000 uses daily, demonstrating significant engagement within the materials science community. The database delivers substantial data volumes—465 terabytes in the last two years alone—equivalent to a vast collection of high-resolution images or movies. This rapid data delivery supports a user base that has grown by 2.5 times since May 2022, highlighting an increasing reliance on its resources. The platform’s impact is further underscored by its over 32,000 citations in peer-reviewed studies, contributing to advancements in fields like battery technology and quantum computing.
Its library contains data on more than 200,000 materials and 577,000 molecules, offering researchers a comprehensive resource. A cloud infrastructure ensures 99.98% uptime, supporting continuous access for scientists globally.
The Materials Project serves as a strong bridge between industry and academia by providing the entire research community with transparently developed open-source tools.
Brian Storey, Toyota Research Institute Vice President
NERSC Supercomputers Enable High-Throughput Data Calculation
NERSC supercomputers are central to the Materials Project’s ability to perform high-throughput calculations on materials. The platform leverages these resources to computationally screen vast libraries of materials, calculating properties using advanced methods verified by experiments. This process facilitates rapid evaluation of numerous materials, significantly speeding up the pace of materials discovery and design. The Materials Project utilizes NERSC for standardized data creation, delivering over 465 terabytes of data in the last two years. This data, including details like electron density, is formatted specifically for machine-learning applications, eliminating months of data preparation. Consequently, researchers can focus on algorithm development and scientific breakthroughs, supported by a cloud infrastructure boasting 99.98% uptime.
Cloud Infrastructure Supports 99.98% Uptime & Data Access
The Materials Project now operates on a cloud-based infrastructure, a move facilitated by partnerships with companies like Amazon Web Services, MongoDB, and Datadog. This system is designed for high availability, currently maintaining 99.98% uptime and supporting rapid data access for its over 650,000 users. This infrastructure enables not only quick searches and large data downloads, but also interactive tools for exploring material relationships. The platform’s ability to consistently operate supports a shift towards remote research, demonstrated during pandemic restrictions, and powers machine-learning applications.
