New research shows that advances in technology could help make future supercomputers far more energy efficient.

Neuromorphic computers are modeled after the structure of the human brain, and researchers are finding that they can tackle difficult mathematical problems at the heart of many scientific and engineering fields.

In a study published in Nature Machine Intelligence, Sandia National Laboratories computational neuroscientists Brad Theilman and Brad Aimone introduce a new algorithm that allows neuromorphic hardware to solve partial differential equations, or PDEs. These equations form the mathematical basis for describing systems such as fluid flow, electromagnetic behavior, and the strength of physical structures.

The results show that neuromorphic systems can not only solve these equations, but can do so with impressive efficiency. According to the researchers, this advance could open the door to the world’s first neuromorphic supercomputer, with major implications for energy-efficient computing in national security and other demanding applications.

A brain-inspired approach to scientific computing

Partial differential equations play a central role in modeling the real world, from forecasting the weather to predicting how materials respond to force. Solving these equations has traditionally required enormous computing power. Neuromorphic computers take a different path, using hardware designs that more closely mirror the way the brain handles information.

“We’re just starting to have computational systems that can exhibit intelligent-like behavior. But they look nothing like the brain, and the amount of resources that they require is ridiculous, frankly,” Theilman said.

For many years, neuromorphic systems were thought to be best suited for pattern recognition or for speeding up artificial neural networks. Few researchers expected them to perform well on mathematically demanding problems like PDEs, which are usually reserved for conventional supercomputers.

Aimone and Theilman, however, see the results as a natural extension of how the brain works. They argue that the brain routinely carries out complex calculations, even though people are rarely aware of them.

“Pick any sort of motor control task — like hitting a tennis ball or swinging a bat at a baseball,” Aimone said. “These are very sophisticated computations. They are exascale-level problems that our brains are capable of doing very cheaply.”

Energy efficiency for national security

The research has important implications for the National Nuclear Security Administration, which is responsible for maintaining the nation’s nuclear deterrence. Across the nuclear weapons complex, supercomputers consume vast amounts of energy to simulate the physics involved in nuclear systems and other critical technologies.

Neuromorphic computing points to a way to cut that energy use while preserving computational capability. By handling PDEs with brain-inspired efficiency, these systems suggest that large and complex simulations could one day be performed using far less power than today’s traditional supercomputers.

“You can solve real physics problems with brain-like computation,” Aimone said. “That’s something you wouldn’t expect because people’s intuition goes the opposite way. And in fact, that intuition is often wrong.”

The researchers said they envision a future where neuromorphic supercomputers play a central role in Sandia’s mission to keep the world safe and secure.

A window into the brain’s secrets

Their research also raises intriguing questions about the nature of intelligence and computation. The algorithm developed by Theilman and Aimone retains strong similarities to the structure and dynamics of cortical networks in the brain.

“We based our circuit on a relatively well-known model in the computational neuroscience world,” Theilman said. “We’ve shown the model has a natural but non-obvious link to PDEs, and that link hasn’t been made until now — 12 years after the model was introduced.”

The researchers believe that neuromorphic computing could help bridge the gap between neuroscience and applied mathematics, offering new insights into how the brain processes information.

“Diseases of the brain could be diseases of computation,” Aimone said. “But we don’t have a solid grasp on how the brain performs computations yet.”

If their hunch is correct, neuromorphic computing could offer clues to better understand and treat neurological conditions like Alzheimer’s and Parkinson’s.

Building the future of computing

While neuromorphic computing is still in its early stages, Sandia’s research is laying the groundwork for transformative advancements. The team hopes their work will inspire collaboration with applied mathematicians, neuroscientists, and engineers to explore the full potential of this technology.

“If we’ve already shown that we can import this relatively basic but fundamental applied math algorithm into neuromorphic — is there a corresponding neuromorphic formulation for even more advanced applied math techniques?” Theilman said.

As Sandia continues to advance neuromorphic computing, the researchers are optimistic about its potential to address some of the world’s most pressing challenges. “We have a foot in the door for understanding the scientific questions, but also we have something that solves a real problem,” Theilman said.

Reference: “Solving sparse finite element problems on neuromorphic hardware” by Bradley H. Theilman, and James B. Aimone, 13 November 2025, Nature Machine Intelligence.
DOI: 10.1038/s42256-025-01143-2

The research was supported by the Department of Energy’s Office of Science through the Advanced Scientific Computing Research and Basic Energy Sciences programs, and by the National Nuclear Security Administration’s Advanced Simulation and Computing program.

Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.