Researchers from Massachusetts Institute of Technology have developed a new chip which is specifically designed to be implemented in neural networks. The chips are energy friendly and can perform powerful Artificial Intelligence (AI) task, enabling future mobile devices to implement “Neutral networks” modeled on the human brain.
The new chip is said to be 10 times as efficient as a mobile GPU and so could enable smart phones run power AI algorithms locally rather than uploading data to the internet for processing.
The GPU has a specialized circuit designed to accelerate the image output in a frame buffer intended for output to a display.
Modern smart phones are equipped with advanced embedded chipsets that can do many different tasks depending on their programming.
GPUs are an essential part of those chipsets and as mobile games are pushing the boundaries of their capabilities, the GPU performance is becoming increasingly important.
The new chip which is being called as “Eyeriss” can help usher in the “Internet of Things” like connected vehicles, appliances, civil engineering structures, manufacturing equipment and even livestock tagging with sensors.
With powerful AI algorithms on board, networked devices could make important decisions locally, entrusting only their conclusions, rather than just pushing raw personal data, to the internet.
MIT researchers presented EYERISS at the “International Solid State Circuits Conference” in San Francisco recently. Researchers used the AI filled Chip to implement a neural network that performance an image recognition task. It was for the first time that a state-of-the-art neural network has been demonstrated on a custom chip.
More details will be updated shortly!