TriMagnetix Develops Energy-Efficient AI Chip Using Nanomagnets

Seattle-based startup TriMagnetix is working on a new generation of computer chips that could dramatically reduce the amount of energy required to power artificial intelligence (AI). As AI adoption continues to surge worldwide, the company believes its nanomagnetic chip technology could offer a more sustainable alternative to today’s semiconductor processors.

Unlike conventional chips that require a continuous flow of electricity to operate, TriMagnetix’s technology relies on short electrical pulses to perform computing tasks. This approach significantly lowers electricity consumption while also producing much less heat, potentially reducing the need for expensive cooling systems in AI data centers.

The startup recently secured $200,000 in funding from climate-focused venture capital firm SNØCAP, providing additional support as it advances development of its first prototype.

AI Growth Is Driving Energy Demand

The rapid expansion of AI services has created unprecedented demand for high-performance chips. At the same time, data centers that power AI applications are consuming increasing amounts of electricity and water, largely due to the cooling requirements of traditional semiconductor hardware.

Industry forecasts suggest that electricity usage at U.S. data centers could more than double over the next decade as AI workloads continue to expand. These growing energy demands have intensified the search for new computing technologies that can deliver better performance while using fewer resources.

The issue has also gained political attention. On Wednesday, the Trump administration announced new initiatives aimed at strengthening U.S. leadership in artificial intelligence, including plans to remove certain environmental regulations that officials argue may slow the construction of new AI data centers.

A Different Approach to Chip Design

According to Madison Hanberry, TriMagnetix’s co-founder and chief of research and development, nanomagnetic computing represents an entirely different way of building processors.

Hanberry explained that the technology faces several challenges before it can become mainstream. These include significant intellectual and financial barriers, as well as the fact that established semiconductor companies have little incentive to move away from existing silicon-based technologies.

Over the years, the semiconductor industry has introduced innovations such as FinFET (Fin Field-Effect Transistor) architecture and Magnetoresistive RAM (MRAM) to improve chip performance. Researchers have also been exploring nanomagnetic computing as a promising long-term solution. A study published last year in the journal Nano Futures highlighted unconventional computing methods, including nanomagnets, and suggested that the field is ready for a clear research roadmap.

From Research Project to Startup

Hanberry’s interest in spintronics—a field that combines electronics with magnetics—began during high school. While studying at Georgia State University, he joined a research laboratory led by Alexander Kozhanov, where he focused on nanomagnetic technology. Kozhanov, now an associate research professor at Duke University, currently serves as an advisor to TriMagnetix.

Hanberry founded the company in 2023, naming it after its unique chip design based on nanomagnetic triangles. His sister, Aspen White, co-founded the startup. The team also includes three software engineers, who have chosen to remain anonymous because they are employed elsewhere.

Building a Prototype

Rather than investing in costly manufacturing facilities, TriMagnetix has partnered with the Washington Nanofabrication Facility at the University of Washington. The collaboration gives the startup access to advanced fabrication equipment and technical expertise needed to build its first working chip.

Hanberry expects the prototype to be completed within six to eight months.

He said many semiconductor startups are discouraged by the enormous infrastructure costs associated with chip manufacturing. TriMagnetix hopes to demonstrate that advanced processors can be developed on a much smaller budget while remaining reliable and scalable.

Investor Sees Major Potential

SNØCAP co-founder and general partner Jonathan Azoff believes the startup’s technology could reshape how future computing systems are designed.

According to Azoff, increasing computing power should not automatically require greater energy consumption. While some technology companies are exploring solutions such as placing nuclear reactors near data centers to meet rising electricity needs, he believes TriMagnetix is pursuing a more fundamental solution by redesigning computing hardware from the ground up.

Beyond AI Data Centers

Although AI infrastructure is a primary target, TriMagnetix sees opportunities in several other industries.

Because nanomagnetic chips are highly resistant to radiation, they could be well suited for aerospace and space applications. In addition, the processors generate far less heat than conventional semiconductor chips, making them attractive for compact wearable devices such as virtual reality (VR) and augmented reality (AR) headsets.

As demand for AI continues to grow, startups like TriMagnetix are exploring new hardware technologies that could help reduce energy consumption while maintaining computing performance. If successful, the company’s nanomagnetic processors could offer a cleaner and more efficient foundation for the next generation of AI-powered systems.

Also read : EU Lawmaker Probing Pegasus Spyware Was Hacked, Researchers Say