New Low-Power AI Platform for Edge Applications
- By John K. Waters
BrainChip, a provider of advanced neuromorphic computing technology, has collaborated with Socionext, which specializes in software-on-a-chip (SoC) solutions for video and imaging systems, to create a low-power platform for AI edge computing applications, the companies announced this week.
Neuromorphic chips of the type BrainChip provides mimic the biological processes of the human nervous system. The company's Akida Neuromorphic System-on-Chip (NSoC) is a third-generation neuromorphic chip inspired by the natural "neuron-spiking" information processing of biological systems. Spiking neural networks (SNN) use even more biologically-realistic models of neurons to carry out computation, and they inherently require less power than traditional convolutional neural networks (CNNs). BrainChip claims its technology "can learn autonomously, evolve, and associate information, just like the human brain."
BrainChip collaborated with Socionext on the implementation of the new platform, which includes the Akida NSoC, to provide a complete ultra-low power AI Edge Network for vision, audio, and smart transducers without the need for a host processor or external memory.
The need for AI in edge computing is growing, and Socionext and BrainChip plan to work together in expanding this business in the global market, the companies said.
"Our neural network technology enables ultra-low power AI technology to be implemented effectively in edge applications," said BrainChip CEO Louis DiNardo, in a statement. "Edge devices have size and power consumption constraints that require a high degree of integration in IC solutions. The combination of BrainChip's technology and Socionext's ASIC expertise fulfills the requirements of edge applications."
In addition to integrating BrainChip's AI technology in an SoC, system developers and OEMs may combine BrainChip's proprietary Akida device and Socionext's processor to create high-speed, high-density, low-power systems to perform image and video analysis, recognition and segmentation in surveillance systems, live-streaming and other video applications, the companies said.
"The Akida family of products allows us to stay at the forefront of the burgeoning AI market," said Socionext's Executive Vice President Noriaki Kub iin a statement. "BrainChip and Socionext have successfully collaborated on the Akida IC development and together, we aim to commercialize this product family and support our increasingly diverse customer base."
BrainChip said it is planning to provide training and technical customer support for its new platform, including network simulation on the Akida Development Environment (ADE), emulation on a Field Programmable Gate Array (FPGA), and engineering support for Akida applications.
John K. Waters is the editor in chief of a number of Converge360.com sites, with a focus on high-end development, AI and future tech. He's been writing about cutting-edge technologies and culture of Silicon Valley for more than two decades, and he's written more than a dozen books. He also co-scripted the documentary film Silicon Valley: A 100 Year Renaissance, which aired on PBS. He can be reached at firstname.lastname@example.org.