Driven by a shift from trial-and-error to predictive design, the Swiss National Centre of Competence in Research MARVEL will close on July 9 after twelve years of reshaping computational science and seeing four materials startups raise more than $800 million in early-stage funding. Launched in 2016, the centre combined quantum-mechanical simulations, computational power, and machine learning, an approach MARVEL director and EPFL professor Nicola Marzari describes as informed by recent industry investment. Marzari notes this reflects a broader move toward designing materials on a computer before physical creation. Over its lifetime, MARVEL researchers predicted and confirmed new quantum materials and strengthened the computational foundations for AI-driven materials discovery, a field now attracting significant global interest.
Quantum Materials Discovery & Predictive Modeling
The impact of the NCCR MARVEL is demonstrated by the fact that four startups have raised more than $800 million in early-stage funding. For much of its history, materials discovery relied on iterative synthesis and testing, but MARVEL actively promoted a paradigm shift. Researchers integrated physics, chemistry, computer science, and experimental data to identify promising materials earlier and understand the principles governing their behavior. This approach yielded key advances, including the prediction and confirmation of new quantum materials exhibiting unusual electronic states relevant to future technologies. These were not isolated incidents; the work transformed longstanding scientific puzzles into clearer design principles. Beyond scientific breakthroughs, MARVEL significantly strengthened the computational foundations of materials science, developing advanced electronic-structure methods and incorporating them into open-source codes used globally. The initiative also developed machine-learning techniques for molecules and materials, creating AI models now widely used. As machine learning evolved from a promising add-on to a central research tool, MARVEL helped prepare the ground for the current surge of interest in AI-driven materials discovery and fostered a Swiss digital ecosystem for materials science built around infrastructures like AiiDA and the Materials Cloud.
Its teams developed advanced electronic-structure methods and incorporated many of these capabilities into open-source codes used by researchers worldwide.
MARVEL’s Integration of Machine Learning & Simulation
Materials science is now characterized by a convergence of computational power, quantum-mechanical simulations, and increasingly sophisticated machine learning algorithms, a combination that was recently largely theoretical but is now rapidly becoming standard practice. Launched in 2016, MARVEL deliberately integrated these disciplines to hasten and systematize materials research. This strategic focus has impacted the scientific and economic environment; four startups have raised more than $800 million in early-stage funding. “Some of the giant internet companies — from Google DeepMind to Microsoft to Meta — have started in the past 2–3 years major efforts in materials design and discovery.” MARVEL’s researchers did not simply apply machine learning as an add-on; they pioneered methods for predicting complex material properties like spectra and electronic structure, laying the groundwork for the current surge in AI-driven discovery.
Beyond the algorithms, MARVEL fostered an open approach, developing open-source codes and a Swiss digital ecosystem, including the AiiDA and AiiDAlab infrastructures and the Materials Cloud platform, to facilitate collaboration and verification. This emphasis on reproducibility and data sharing has created a national platform for computational materials research, ensuring that simulations can be easily run, compared, and built upon by the global scientific community. The initiative’s legacy extends beyond specific discoveries to a fundamentally altered approach to materials innovation.
Some of the giant internet companies – from Google DeepMind to Microsoft to Meta – have started in the past 2-3 years major efforts in materials design and discovery, and this is mirrored by an influx of startups in the field – just four of them raising more than 800 million dollars in early-stage funding.
Advancing Computational Infrastructure: AiiDA & Materials Cloud
The shift toward predictive materials design increasingly relies on robust computational infrastructure, a need the NCCR MARVEL addressed by developing the AiiDA and Materials Cloud platforms. This focus on reproducibility and open science represents a significant departure from the historically trial-and-error approach to materials discovery, where synthesis and testing often dominated the process. MARVEL’s digital ecosystem is not merely a repository for data; it actively facilitates connections between simulations, data handling, and experiments, even steering automated experimental platforms in some instances. The impact extends beyond academia, with MARVEL fostering ties with industry partners in sectors ranging from energy to pharmaceuticals. This collaboration resulted in a move from specialist software toward practical tools applicable in industrial settings, a key factor reflected by an influx of startups in the field, four of which have raised more than $800 million in early-stage funding. The culmination of these efforts will be celebrated on July 9 in Lausanne, marking the official closing event for the 12-year NCCR MARVEL initiative and solidifying its legacy as a pioneer in AI-driven materials science.
Societal Impact: Materials for Energy & Manufacturing
The impact of the NCCR MARVEL extends beyond academic publications, demonstrably influencing the landscape of materials-based innovation and attracting significant investment. A surge in startup funding, with four companies raising more than $800 million in early-stage capital, illustrates the economic momentum correlated with MARVEL’s approach to materials discovery. This influx of capital signals growing confidence in computationally designed materials as viable commercial ventures, particularly in sectors demanding advanced performance characteristics. The initiative’s focus on practical applications is evident in its research portfolio, encompassing materials critical for renewable energy, advanced batteries, and high-performance manufacturing. Researchers did not solely pursue fundamental science; they actively investigated materials for solar cells, water splitting, and solid-state ion conductors, addressing challenges with direct societal relevance.
This translational focus, combined with a commitment to open science, fostered collaboration with a diverse industrial community, including companies like BASF and Stellantis. “A key achievement was the move from software built primarily for specialists toward practical tools that can be adopted in industrial settings,” reflecting a deliberate effort to bridge the gap between research and real-world implementation. As MARVEL concludes its 12-year run with a closing event on July 9, its legacy lies in establishing a digital ecosystem for materials science, with tools like AiiDA and the Materials Cloud serving not merely as research aids but as platforms designed for reproducibility, data sharing, and accessibility, ensuring that computational workflows can be integrated with automated experimental platforms.
