48 core neuromorphic AI chip uses resistive memory

Find the Latest SoC Solutions
for... Automotive... IoT... Security... Audio... Video

eenewseurope.com, Aug. 17, 2022 – 

A team of researchers in the US and China has designed and built a neuromorphic AI chip using resistive RAM, also known as memristors.

The 48 core NeuRRAM chip developed at the University of California San Diego is twice as energy efficient as other compute-in-memory chips and provides results that are just as accurate as conventional digital chips.

Computation with RRAM chips is not necessarily new, and many startups and research groups are working on the technology. However it generally leads to a decrease in the accuracy of the computations performed on the chip and a lack of flexibility in the chip's architecture.

The NeuRRAM chip is also highly versatile and supports many different neural network models and architectures. As a result, the chip can be used for many different applications, including image recognition and reconstruction as well as voice recognition.

"The conventional wisdom is that the higher efficiency of compute-in-memory is at the cost of versatility, but our NeuRRAM chip obtains efficiency while not sacrificing versatility," said Weier Wan, the paper's first corresponding author and a recent Ph.D. graduate of Stanford University who worked on the chip while at UC San Diego, where he was co-advised by Gert Cauwenberghs in the Department of Bioengineering. "It's the equivalent of doing an eight-hour commute for a two-hour work day."

RRAM and other emerging memory technologies used as synapse arrays for neuromorphic computing were pioneered in the lab of Philip Wong, Wan's advisor at Stanford and a main contributor to the development, along with researchers at Standord in the US and Tsinghua University in China.

click here to read more

 Back