AI Object Recognition Accelerates Neuromorphic Computing

BrainChip’s hardware acceleration board, simply dubbed BrainChip Accelerator, is an eight-lane, PCI-Express add-in card that increases the speed and accuracy of the object recognition function of BrainChip Studio software by up to six times. It simultaneously increases video channels of the host system to 16 per card. Described as consuming very low-power, the card easily installs within existing video surveillance systems without the need to upgrade power systems or thermal management.


Tailored to law enforcement and intelligence organizations, BrainChip Studio software rapidly identifies objects in large amounts of archived or live streaming video. Processing multiple video streams simultaneously, the accelerator add-in card enables users to search increasing amounts of video faster, with a higher probability of object recognition and lower total cost of ownership. The system learns from a single low-resolution image, which can be as small as 20 x 20 pixels. The card is also said to excel in low-light, low-resolution, and noisy environments.

Free Monthly Newsletter

Compelling read? Subscribe to FierceEmbeddedTech!

The embedded tech sector runs the market’s trends. FierceEmbeddedTech subscribers rely on our suite of newsletters as their must-read source for the latest news, developments and analysis impacting their world. Sign up today to get news and updates delivered to your inbox and read on the go.
BrainChip Accelerator

BrainChip Accelerator embarks as the first commercial implementation of a hardware-accelerated spiking neural network system, making significant progress in the development of neuromorphic computing, a leg of artificial intelligence that simulates neuron functions. Processing is done by six BrainChip Accelerator cores in a Xilinx Kintex Ultrascale field-programmable gate array (FPGA). Each core performs user-defined image scaling, spike generation, and spiking neural network comparison to recognize objects. In terms of power, each core consumes approximately 1W while processing up to 100 fps. In comparison to GPU-accelerated deep learning classification neural networks like GoogleNet and AlexNet, this is reported as a seven times improvement of frames/second/watt.


BrainChip Accelerator is compatible with Windows or Linux operating systems and is currently available to select law enforcement and intelligence agencies as an integrated server appliance. For further details, contact BrainChip Holdings Ltd.

Suggested Articles

U.S. trade groups SEMI and SIA say semi workers need to make on-site visits to chip plants abroad for installation and repairs

IC Insights analyst sees economic softness that began in 2019 hitting analog sales in many regions again in 2020.

Omdia forecast a $1.7B reduction in revenues for auto power semiconductors