BREAKING
Discovered Materials offers $1 million prize to solve materials science problems
The startup founded by IIT Madras alumni aims to benchmark AI capabilities while tackling critical bottlenecks in semiconductor development.
San Francisco-based deep-tech startup Discovered Materials has launched a $1 million prize challenge to solve five longstanding problems in materials science, aiming to test how far artificial intelligence and researchers can push scientific discovery. Co-founded by IIT Madras alumni Advaith Sridhar and Akash Ramdas, the company worked alongside academic researchers from leading universities to identify key scientific bottlenecks that have remained unsolved despite years of study.
The competition is open to eligible individuals, teams, and organisations. There is no fixed deadline for submissions; instead, cash prizes will be awarded whenever a proposed solution satisfies the specific scientific criteria established for each challenge.
Why it matters
For founders, operators, and investors in hardware and deep tech, the initiative targets core bottlenecks in semiconductor development and advanced physics. Progress in electronic devices, energy-related technologies, and microchip manufacturing depends on understanding complex material properties and quantum behavior. If researchers or autonomous systems resolve these questions, it could lead to faster quantum chemistry modeling, improved semiconductor design, and better thermal management in microprocessors.
Beyond the scientific solutions themselves, the challenge creates an open testbed to determine whether advanced AI models can drive genuine empirical discovery or if current automated systems hit a wall when confronted with unsolved theoretical physics.
The five scientific challenges
The five problems chosen for the competition address hard theoretical and practical questions across materials science. The largest individual reward is earmarked for solving cuprate superconductivity, while four additional awards target interface electrical properties, quantum calculations, light interactions, and thermal transport.
| Challenge Area | Scientific Objective | Prize Amount |
|---|---|---|
| Cuprate Superconductivity | Understanding high-temperature cuprate physics | $400,000 |
| Semiconductor Interfaces | Predicting electrical properties across solid interfaces | $200,000 |
| Quantum Chemistry | Finding faster and accurate methods for molecular calculations | $200,000 |
| Light-Material Interaction | Understanding how light alters material behaviour | $100,000 |
| Thermal Transport | Predicting heat movement across solid interfaces | $100,000 |
Finding answers to these problems carries direct practical value for the technology sector. For example, predicting how heat travels across solid material boundaries is essential to managing thermal limits in modern computing chips. Similarly, understanding light-material interactions and the electrical properties of semiconductor interfaces can directly improve chip manufacturing techniques.
AI agents fail internal test
Before launching the public challenge, Discovered Materials conducted an internal trial to evaluate current AI capabilities against the five problems. The startup deployed advanced AI models configured to run as autonomous agents, granting them access to the internet and academic research databases.
The internal test showed the limitations of current automated research tools. Even after working independently for several days, the AI agents made little progress toward solving any of the five challenges, according to reports.
"How do we know AI is making progress?" Sridhar said, highlighting the difficulty of measuring whether machine learning systems are generating new scientific knowledge rather than merely synthesising existing literature.
Stanford University professor Felipe Jornada, who assisted in formulating the challenge problems, noted that theoretical understanding in the sector continues to lag behind experimental observations. "Across materials science, there are phenomena we observe but cannot explain," Jornada said.
Funding and company background
Sridhar and Ramdas met more than a decade ago while studying at IIT Madras. They later established Discovered Materials to build AI-powered software designed to accelerate the discovery of new materials for semiconductor chips.
| Parameter | Details |
|---|---|
| Founders | Advaith Sridhar, Akash Ramdas |
| Alma Mater | IIT Madras |
| Headquarters | San Francisco, California |
| Accelerator Batch | Y Combinator Spring 2026 |
| Seed Funding | $9 million |
| Seed Round Month | August 2026 |
| Core Product Focus | AI-powered semiconductor material discovery |
The startup participated in Y Combinator's spring 2026 accelerator batch. In August 2026, Discovered Materials raised $9 million in seed funding to expand its research into semiconductor materials. The company has not disclosed its valuation or the names of the investors participating in the seed round.
What happens next
The challenge remains open until solutions meeting the scientific criteria are submitted and verified. Discovered Materials and its academic collaborators plan to review entries as they arrive, using the submission pool to track whether human scientists or evolving AI agents make the first breakthroughs on the five problems.