The U.S. Department of Energy has struck a $1 billion deal with Advanced Micro Systems (AMD) to build two supercomputers that have unprecedented power to supercharge scientific advances ranging from nuclear power to developing cancer treatments.
The partnership, first reported by Reuters on Monday, will ensure the U.S. government has the necessary computing power to accommodate enormous amounts of data—and could deliver about three times the AI capacity of current supercomputers. The artificial intelligence-powered supercomputers could be deployed for advancing nuclear power and replicating fusion—the process that fuels the sun and creates massive amounts of energy.
“We’re going to get just massively faster progress using the computation from these AI systems that I believe will have practical pathways to harness fusion energy in the next two or three years,” Energy Secretary Chris Wright told Reuters.
Shares of AMD rose nearly 1% in midday trading on Monday. The semiconductor maker’s stock has more than doubled in value this year, and the company recently forged a partnership to supply its chips to OpenAI to build out AI infrastructure.
Neither the Department of Energy nor AMD responded immediately to a request for comment from Fast Company.
SUPERCOMPUTER TIMELINES
The forthcoming supercomputers could be used to further advancements in technologies for defense and national security, including helping the U.S. government to manage its arsenal of nuclear weapons, along with accelerating drug discovery for the treatment of cancers, Wright told Reuters. “My hope is, in the next five or eight years, we will turn most cancers, many of which today are ultimate death sentences, into manageable conditions.”
The first of two supercomputers, Lux, could be up and running within the next six months and will be based on AMD’s MI355X artificial intelligence chips and the company’s central processing units (CPUs) and networking chips. The supercomputer system is co-developed by AMD, Hewlett Packard Enterprise, Oracle Cloud Infrastructure, and Oak Ridge National Laboratory (ORNL).
The timeline for Lux will mark the fastest deployment of this size of supercomputer that AMD’s CEO, Lisa Su, has ever seen, she told Reuters. “This is the speed and agility that we wanted to [do] this for the U.S. AI efforts.”
A second, even more advanced supercomputer, called Discovery, will have a longer timeline for completion and is expected to be ready for operations sometime in 2029.