Jason Eshraghian and Rui-Jie Zhu built an AI that runs on the power of a lightbulb.
Santa Cruz, CA, USA · Jason Eshraghian, Rui-Jie Zhu
Published July 15, 2026
Large language models are notorious for their energy appetite. A UC Santa Cruz professor and his graduate student found a way to strip the most expensive step out of the math, and ran a billion-parameter-scale model on 13 watts, about what it takes to light a room.
The story
The person and the place
Jason Eshraghian, assistant professor of electrical and computer engineering at UC Santa Cruz's Baskin School of Engineering, and Rui-Jie Zhu, his graduate student and the paper's first author.
The problem
Running large language models costs real money and real energy, and Eshraghian wanted to know whether that cost was actually necessary. "Language models like ChatGPT are incredibly expensive in terms of resources," he said.
The moment they didn't wait
Instead of accepting that bigger, pricier hardware was the only path forward, Eshraghian and Zhu set out to eliminate matrix multiplication, the single most computationally expensive step in running a neural network, and then built custom hardware to prove it could be done. "We got the same performance at way less cost — all we had to do was fundamentally change how neural networks work. Then we took it a step further and built custom hardware," Eshraghian said. They built the low-power prototype system in three weeks. As he put it: "We are just a small academic lab...capable of competing with the giants."
The custom hardware ran a billion-parameter-scale language model on just 13 watts of power, more than 50 times more efficient than typical GPU hardware. On standard GPUs, the same approach cut memory use tenfold and sped things up 25%, while holding performance comparable to state-of-the-art transformer models.
"We are just a small academic lab...capable of competing with the giants." — Jason Eshraghian, Santa Cruz Sentinel via GovTech Insider
Verified sources
Sources
Every claim in this story is checked against primary sources. Verify it yourself.