manufacturer of AI chips With the goal of accelerating medical and genetic advancements, Nvidia announced on Wednesday that it and its research partners had developed what they describe as the largest artificial intelligence system to date for biological research.
Evo 2, the new artificial intelligence system, is able to read and create genetic code from all living species.
Researchers expect that by identifying patterns in massive volumes of data that would typically take years to evaluate manually, AI technology can significantly speed up research.
The algorithm was trained using about 9 trillion genetic data points from more than 128,000 distinct creatures, including humans, plants, and bacteria.
It correctly detected 90% of potentially dangerous mutations in the breast cancer gene BRCA1 in early testing.
According to researchers, this could aid in the creation of more specialized medications, such as gene therapies that exclusively target particular cells.
2,000 Nvidia H100 CPUs on Amazon’s cloud infrastructure were used to build the model.
Evo 2, created in collaboration with Stanford University and the Arc Institute, is now openly accessible to researchers everywhere via Nvidia’s BioNeMo research platform.
According to Stanford University associate professor Brian Hie, “creating new biology has historically been a time-consuming, uncertain, and artisanal process.”
With Evo 2, we enable researchers to better understand the biological architecture of complex systems.
Beyond medicine, scientists think the technique could aid in the development of new methods for decomposing pollution and crops that are more resilient to climate change.
The initiative combines the computational capacity of Nvidia with the $650 million-funded nonprofit research organization Arc Institute, which was established in 2021.
To address long-term scientific issues, the institute collaborates closely with Stanford, UC Berkeley, and UC San Francisco.
