The Genesis Mission brings together DOE national laboratories, industry, academia and other research organisations
AMD has announced its role in advancing the U.S. Department of Energy’s (DOE) Genesis Mission through the next generation of AI and high-performance computing (HPC) systems, with the upcoming Lux supercomputer expected to become the mission’s first operational platform and Discovery planned as a future exascale flagship system.
Lux is scheduled to begin supporting funded research projects in October 2026. The system is designed to combine AI, simulation, data analytics and scientific computing on a single platform, helping accelerate research in energy, materials science, medicine, national security and advanced manufacturing.
The Genesis Mission brings together DOE national laboratories, industry, academia and other research organisations to use AI for scientific and engineering breakthroughs. By integrating simulation, data and AI workflows, the initiative aims to shorten the time required for discovery and improve the productivity of U.S. research programs.
According to AMD, Lux will provide researchers with an open and interoperable computing environment capable of supporting traditional HPC applications alongside cloud-native AI tools.
Thomas Zacharia, senior vice president of Strategic Technology Partnerships and Public Policy at AMD, said the company is helping turn the Genesis Mission into a working scientific platform. He added that Lux will enable researchers to combine simulation, AI, data and emerging computing workflows on a high-performance foundation, while Discovery will extend that foundation for the next generation of American science.
The Lux architecture will include AMD Instinct MI355X GPUs, AMD EPYC CPUs, AMD Pensando networking technologies and an open software ecosystem. The system is being designed to support the entire scientific workflow, from instrument data acquisition to AI-driven insight.
AMD said Lux is also intended to support future capabilities such as the American Science Cloud, collaborative AI model development, autonomous laboratories, digital twins, persistent AI services and workflows that span CPUs, GPUs and quantum processors.
The company pointed to the existing Frontier supercomputer, also powered by AMD technology, as an example of how converged AI and HPC can accelerate scientific research. Frontier is being used for applications including rapid plant-imaging analysis and the study of materials for tritium production in future fusion reactors.
These types of workflows are expected to expand with Lux and later Discovery, with research opportunities covering critical mineral recovery, fusion energy, hydropower, flood response and electrical grid resilience.
Discovery, which is planned in collaboration with the DOE, Oak Ridge National Laboratory and HPE, is intended to become a long-term platform for converged AI and exascale computing.
The future system is expected to use 6th Gen AMD EPYC processors, AMD Instinct MI430X GPUs, AMD Pensando networking technologies and the ROCm software stack. AMD said Discovery is being designed to support large-scale simulations, scientific models and integrated discovery workflows with native FP64 precision for high-accuracy computing.
The company described Discovery as the next step beyond Frontier and Lux, providing a leadership platform for deeper convergence between AI and high-performance computing in U.S. scientific research.
