Trillion Parameter Consortium

The Trillion Parameter Consortium (TPC) is a global initiative that brings together scientists from government laboratories, academia, research institutes, and industry to tackle the immense challenges of building large-scale AI systems for scientific discovery. By focusing on the development of massive generative AI models, the consortium aims to advance trustworthy and reliable AI tools that address complex scientific and engineering problems. 

TPC’s mission includes creating scalable model architectures, organizing and curating diverse scientific datasets, optimizing AI technologies for exascale computing platforms, and developing evaluation methods to ensure reliability, trustworthiness, and progress in scientific task learning. By fostering collaboration, TPC seeks to accelerate innovation and overcome barriers to applying AI in science at an unprecedented scale. 

At the core of TPC’s vision is the creation of a global network of researchers and resources to support the development of next-generation AI systems. This includes preparing high-quality datasets tailored to specific scientific domains, designing and evaluating state-of-the-art model architectures, and ensuring these systems meet high standards of bias mitigation, trustworthiness, and goal alignment. Recognizing the monumental computational demands of training trillion-parameter AI models, TPC leverages exascale computing platforms provided by partners like U.S. Department of Energy labs and international institutions. 

A true community effort, the founding partners of TPC include Argonne, Barcelona Supercomputing Center, RIKEN and others and has grown to over 80 participating organizations. Through shared expertise, tools, and workflows, TPC promotes collaboration across disciplines and institutions, enabling scientists to harness AI’s transformative potential while addressing challenges too vast for any single organization to manage alone.