$1 Billion U.S.–AMD Deal to Build Next-Generation Supercomputers

U.S. AMD

The United States has entered a $1 billion partnership with Advanced Micro Devices (AMD) to construct two world-class supercomputers designed to address some of the nation’s most critical scientific and security challenges. According to officials, the systems will advance computing power for projects ranging from nuclear reactor modeling to cancer treatment research and national defense analysis.

---

AMD’s Role in the Federal Supercomputing Push

The deal marks a major milestone for AMD, solidifying its place as a core partner in America’s high-performance computing (HPC) strategy. The company’s cutting-edge EPYC processors and Instinct AI accelerators will serve as the engine of the new supercomputers, offering an estimated performance measured in exaflops — a trillion trillion calculations per second.

The systems are expected to rank among the most powerful supercomputers in the world, placing the U.S. at the forefront of global computational science once again. AMD executives said the partnership underscores the importance of public-private collaboration in keeping America’s technology infrastructure competitive.

“This collaboration is about building computing capacity that matches our scientific ambitions,” an AMD spokesperson noted.

---

Scientific Goals of the Supercomputer Project

The two machines will focus on a range of high-impact scientific domains:

• • Clean Energy Research: Simulating nuclear fusion and reactor designs to develop safer and more efficient energy sources.

  • Medical Advancement: Accelerating drug discovery and cancer treatment research through AI-driven molecular modeling.

• • Climate and Earth Sciences: Running large-scale climate models to improve extreme weather forecasting and disaster readiness.

  • National Security: Providing the Department of Energy (DOE) and defense agencies with data-intensive capabilities for cyber resilience and nuclear safety.

Each supercomputer will operate under DOE oversight and integrate into the national laboratory network, allowing scientists and researchers to run petabyte-scale simulations on shared federal resources.

---

Where the Systems Will Be Built

Preliminary plans indicate that one system will be installed at Oak Ridge National Laboratory in Tennessee, home to the existing Frontier supercomputer, while the second is expected to go to Lawrence Livermore National Laboratory in California. Both facilities have longstanding partnerships with AMD and the DOE in developing AI and HPC systems.

Officials expect the project to generate hundreds of engineering and manufacturing jobs across AMD’s supply chain, particularly in the semiconductor hubs of Texas and California.

---

Strengthening America’s Computational Leadership

The supercomputer initiative arrives amid intense global competition in HPC, as China and Europe continue to invest heavily in their own exascale programs. U.S. officials say this partnership is critical to maintaining technological leadership and ensuring that future research and national security operations rely on domestic hardware and innovation.

“These machines aren’t just computers; they’re national assets,” a senior DOE administrator said. “They define how fast we can solve the hardest scientific problems.”