Trending:

US is building massive supercomputers to tackle problems — from nuclear energy to cancer treatments

FP News Desk October 28, 2025, 09:02:28 IST

The United States has partnered with AMD in a $1 billion initiative to build two supercomputers aimed at advancing research in energy, security, and medicine

Advertisement
The US is developing the two supercomputers to ensure it has sufficient computing power to support increasingly complex experiments that demand massive data-processing capabilities.
The US is developing the two supercomputers to ensure it has sufficient computing power to support increasingly complex experiments that demand massive data-processing capabilities.

The United States has entered a $1 billion partnership with Advanced Micro Devices (AMD) to construct two powerful supercomputers designed to address complex scientific challenges, from nuclear power and cancer treatment to national security. The agreement was announced by Energy Secretary Chris Wright and AMD Chief Executive Officer Lisa Su in an interview with Reuters.

The two systems will strengthen the country’s capacity to run increasingly intricate experiments that depend on vast data processing capabilities. Wright said the machines would “supercharge” innovation in nuclear and fusion energy, defence technologies, and drug development. He noted that replicating fusion, the reaction that powers the sun, requires creating extreme conditions by compressing light atoms in plasma under intense heat and pressure. “We’ve made great progress, but plasmas are unstable, and we need to recreate the centre of the sun on Earth,” he explained.

STORY CONTINUES BELOW THIS AD

Wright added that the new computing power could significantly accelerate progress in harnessing fusion energy within the next few years. The systems will also support the management of the US nuclear weapons arsenal and enable molecular-level simulations for cancer treatment research. “My hope is that in the next five or eight years, we will turn most cancers, many of which today are ultimate death sentences, into manageable conditions,” he said.

Two new machines: Lux and Discovery

The first system, named Lux, is scheduled to be built and operational within six months. It will use AMD’s MI355X artificial intelligence chips, along with AMD central processors and networking components. The project is being developed jointly by AMD, Hewlett Packard Enterprise, Oracle Cloud Infrastructure, and Oak Ridge National Laboratory (ORNL). Su described the rollout as the fastest deployment of its kind, saying, “This is the speed and agility that we wanted to achieve for the US AI efforts.”

According to ORNL Director Stephen Streiffer, the Lux supercomputer will provide nearly three times the AI capacity of current systems. The second, more advanced machine, Discovery, will be built around AMD’s MI430 series of AI chips, which integrate high-performance computing features with AI processing capabilities. Designed by ORNL, HPE, and AMD, Discovery is expected to be delivered in 2028 and become operational in 2029. Streiffer said he anticipated major gains in performance but declined to estimate its final capacity.

The Department of Energy will host the supercomputers, while private partners will supply the hardware and investment, sharing the resulting computing resources. A DOE official said the AMD-based systems mark the beginning of a broader series of collaborations between the department and private industry to develop next-generation computing infrastructure across the country.

(With agency inputs)

Home Video Shorts Live TV