Establishment of TPC Alliance: target trillions of parameter AI model to promote scientific discovery
CTOnews.com November 16 news, the industry's leading scientific research institutions, the United States National Supercomputing Center and many leading companies in the AI field, recently jointly formed the Trillion Parameter Consortium (TPC).
CTOnews.com learned from the report that the TPC Alliance is composed of scientists from laboratories, research institutions, academia and industry around the world, with the goal of jointly advancing AI models for scientific discovery, with a special focus on giant models with one trillion or more parameters.
The TPC Alliance is already developing scalable model architectures and training strategies, organizing and collating scientific data for model training, and optimizing AI libraries for current and future terahertz subcomputing platforms.
TPC aims to create an open community of researchers to develop large-scale generative AI models for scientific and engineering problems, and in particular, will initiate joint projects to avoid duplication of effort and share methods, approaches, tools, knowledge and workflows. In this way, the consortium hopes to maximize the impact of these projects on the broader AI and scientific communities.
TPC also aims to create a global network of resources, data and expertise. Since its inception, the consortium has established numerous working groups to address the complexities of building large-scale AI models.
The terahertz of secondary computing resources required for training will be provided by several national laboratories of the U.S. Department of Energy (DOE) and several TPC founding partners in Japan, Europe and other countries. Even with these resources, training takes months.
Rick Stevens, associate director of computing, environmental and life sciences at the U.S. Department of Energy's Argonne National Laboratory and professor of computer science at the University of Chicago, said: "In our laboratories and a growing number of partner institutions around the world, teams are beginning to develop cutting-edge AI models for scientific use and are preparing large amounts of previously undeveloped scientific data for training.
