Nvidia’s chips run the data centers that train the world’s largest models. On September 23 the company said it wants those same systems to help discover drugs.
The chipmaker announced a new research group in Israel, called Applied AI Architecture, aimed at using artificial intelligence to speed up medical and scientific breakthroughs. The unit will sit inside Nvidia’s global chief technology officer’s office, run by Michael Kagan, the Mellanox co-founder who is the company’s most senior Israeli executive and reports directly to chief executive Jensen Huang.
The group is led by Nati Daniel, who holds a doctorate in artificial intelligence from the Technion and has held senior positions at Intel, GE and Huawei. Nvidia plans to recruit dozens of researchers over the coming year, adding to the handful already working at its Tel Aviv offices.
The mandate is deliberately long-term. Rather than build products for immediate sale, the researchers will explore how AI can become a practical engine for scientific discovery, with the work meant to mature into technology that eventually feeds Nvidia’s systems.
The first projects show the ambition. The group is working with Sheba Medical Center and Mount Sinai in New York on a genomic foundation model designed to help researchers study the roughly 98 percent of the human genome that remains poorly understood. The goal is better disease prevention, sharper diagnostics and more personalized medicine.
It is also part of Nvidia’s global collaboration with Eli Lilly. The two companies plan to invest about $1 billion over five years to apply AI to drug discovery, combining genomic and sequencing data with chemistry and biology in an effort to shorten the path from basic research to new treatments.
A third project, called GENMAT, began months ago with a grant of about 6 million euros from the European Union’s Horizon program. It brings companies, universities and research institutes together to develop AI tools that can accelerate the discovery of new materials, including those for aviation and automotive use.
The push builds on work Nvidia has pursued for years. The company has offered tools for genomics and drug discovery, and has courted researchers who see the human body as the next large dataset. It has partnered with hospitals, universities and drugmakers, positioning itself less as a chip supplier and more as a platform for the field’s most ambitious research.
Israel was chosen in part for its concentration of AI and medical research. Sheba, for instance, recently became the first hospital outside the United States to sign a collaboration agreement with OpenAI. Nvidia is said to be developing further partnerships in the country and is expected to announce more in the coming months.
The move deepens a relationship that began when Nvidia bought Mellanox in 2020. The Israeli development center has since become one of the company’s most important research hubs outside the United States, and the workforce there has grown steadily. Kagan’s placement atop the new group signals how central the country has become to Nvidia’s research plans.
The announcement landed on a day when the company’s stock was drawing different attention. Bloomberg reported that Nvidia’s sinking valuation is sending a warning signal about its ability to keep growing profits at the pace investors have come to expect. Huang has pushed back, calling Nvidia the world’s first and only growth value stock.
The two stories sit awkwardly together. On one side, Nvidia is planting a flag in a market, healthcare and drug discovery, that could keep its accelerators busy for decades. On the other, the market is questioning how much longer the current run can hold.
Medical AI is one of Nvidia’s answers. The company has argued that genomics, imaging and drug design will eventually consume vast amounts of its chips, opening a second demand stream beyond the cloud providers that dominate today’s sales. A research group in Israel is a cheap way to buy a claim on that future while the commercial market matures.
The bet is not without risk. Drug discovery has absorbed billions in AI spending with few approved drugs to show for it. Nvidia’s group is studying concepts that may take years to become usable technology, and the payoff, if it comes, will arrive on a timeline that quarterly earnings cannot capture.
For now, the Israel group is a small, deliberate signal. It is tiny by the standards of a company whose market value is measured in trillions, but its placement under Kagan gives it visibility that signals intent. Nvidia is telling the market it sees a second act beyond training chatbots, one built on laboratories and hospitals as much as data centers. The research bets cost little against a balance sheet of Nvidia’s size, but they carry a message that the company’s growth cannot rest on data centers alone. Whether the stock believes it is a separate question.
