Scientists built a chip that works in temperatures hotter than lava

Date:

Share:


Heat has been the ceiling that has stopped us from building electronic devices that survive in harsh conditions. Not only do electronic devices begin to thermally throttle at high temperatures, but if they are pushed beyond 200 degrees, they stop working entirely.

But that might change, as a team at the University of Southern California may have broken through the thermal ceiling. In a study published in Science, researchers developed a new memory chip that kept working reliably at 700 degrees Celsius. For reference, that’s hotter than molten lava.

The device showed no signs of giving up. Seven hundred degrees was simply the limit of their testing equipment. “You may call it a revolution,” said Joshua Yang, the lead researcher on the project. “It is the best high-temperature memory ever demonstrated.”

How did they pull this off?

The chip is built like a tiny sandwich. Tungsten on top, a thin ceramic layer in the middle, and graphene on the bottom. Each of these materials handles extreme heat well on its own, but the magic happens at the graphene layer.

In a normal chip, heat pushes metal atoms through the ceramic layer until both sides connect, resulting in a short circuit. Graphene prevents that from happening. According to Yang, Graphene and Tungsten act as oil and water. Tungsten atoms that drift toward it simply cannot grip the surface and back away. Since there’s no anchor, the chip doesn’t short-circuit.

The best part? The team stumbled onto this entirely by accident. “To be honest, it was by accident, as most discoveries are,” Yang said.

Why breaking the thermal ceiling matters?

The thermal ceiling has ensured that we cannot make machines that can handle extreme temperatures. This means we have not been able to send probes to planets with extremely high temperatures or conduct any deep-earth drilling experiments. The new chip can allow us to conduct such experiments. 

Another use of this chip is in AI. More than 90 percent of the computing in AI systems, like ChatGPT, involves a single type of mathematical calculation. This chip can perform those calculations the instant electricity passes through it, making it far faster and more energy efficient than anything we use today.

Note that a finished product is still years away. But as Yang put it, “The missing component has been made.” Once it comes into production, it will solve one of the biggest problems hindering the advancement in chip technology.



Source link

━ more like this

Apple pulled the most un-Apple move with a price drop on the Studio Display XDR

Just weeks after launch, Apple has quietly done something it rarely does this quickly: it’s cut the price of its brand-new Studio Display...

AI Models Lie, Cheat, and Steal to Protect Other Models From Being Deleted

In a recent experiment, researchers at UC Berkeley and UC Santa Cruz asked Google’s artificial intelligence model Gemini 3 to help clear up...

‘Thank You For Generating With Us!’ Hollywood’s AI Acolytes Stay on the Hype Train

While this type of hype is predictable at industry-led events, again and again summit attendees were reminded that generative AI isn’t just another...

Android 17 is stepping up location privacy in a big way

For years, location permissions have been a bit of a mess on Android. You open an app, it asks for your location, and...

Mr. Resident Evil signs a deal with Mr. Stellar Blade

Resident Evil legend Shinji Mikami's new studio, Unbound Inc., has been acquired by Shift Up, the company behind Stellar Blade and Goddess of...
spot_img