EPFL engineers develop gallium nitride transistor at EPFL; reduces energy loss in high-voltage microchips

EPFL engineers develop gallium nitride transistor at EPFL; reduces energy loss in high-voltage microchips

News - EPFL

An AI capable of doubt can optimize scientific discovery

A team at EPFL have developed a framework that trains Large Language Models to create AI systems that can find the best possible setup or “recipe” for a scientific experiment.

New transistor brings high voltage to microchip scale

By harnessing the properties of gallium nitride, EPFL engineers have built a tiny transistor that can handle very high voltage with minimal energy loss, a key requirement for AI data centers, electric vehicles, and solar power systems.

What triggered the devastating flash flood in Nepal?

Published: 28.08.26 — A raging flood wreaked havoc in a valley in Nepal near the border with China. We spoke with EPFL professor Christophe Ancey, an expert in natural disasters, about the underlying mechanisms and whether such an event could happen in Switzerland.

Published: 27.08.26 — Robots are often promoted for their increasing intelligence and autonomy. Yet a growing body of research at EPFL suggests that a robot’s contribution to sustainability goals may be an equally important measure of success.

New agreement protects EPFL’s built environment and keeps it dynamic

Published: 24.08.26 — EPFL’s very first buildings and their surrounding areas were just classified as historical monuments. In connection with this, the School has entered into an agreement with Vaud Canton and the ETH Board to preserve its emblematic 20th-century architecture while enabling the Lausanne campus to keep changing with the times.

Retreating glaciers are taking entire worlds with them

Published: 21.08.26 — Some 40% of Switzerland’s glacial reserves have disappeared in the past 25 years. That means the country is losing not just a water and energy resource but also living ecosystems that we’re only now discovering. Two EPFL researchers describe the scope of the issue.

How attackers persuade AI agents to break the rules

Published: 19.08.26 — As AI assistants evolve into AI agents, a new EPFL study has found that the biggest safety risks in their use may not come from single malicious prompts, but from carefully orchestrated conversations.

Import & publish the news

Are you sure you want to import this news into ?

This news will be sent to its subscribers.