New IBM Programming System for Big Data Modeled on Human Brain
IBM today announced a new computer programming system based on its cutting-edge silicon chips modeled on characteristics of the human brain.
The new software ecosystem is needed to program the "neurosynaptic chips" because the traditional computing architecture used for the past 50 years can't handle today's Big Data revolution, IBM said. That Von Neumann architecture can process huge amounts of data, but can't think--a characteristic of the human brain that the company is trying to emulate.
"With the advent of Big Data, which grows larger, faster and more diverse by the day, this [Von Neumann] type of computing model is inadequate to process and make sense of the volumes of information that people and organizations need to deal with," IBM stated in its description of the SyNAPSE project that developed the chips.
The new chip architecture was inspired by human brain characteristics such as cognitive functionality, low power and small volume, and could result in breakthroughs such as "a new generation of intelligent sensor networks that mimic the brain's abilities for perception, action and cognition," the company said.
The new programming model takes advantage of several breakthroughs developed by IBM scientists, including: a multithreaded, massively parallel and scalable simulator; a neuron model; a programming model based on "corelets," which are composable, reusable programming components; a cognitive system store library; and a laboratory described as "a novel teaching curriculum that spans the architecture, neuron specification, chip simulator, programming language, application library and prototype design models."
The company said the new software ecosystem addresses all parts of the computer programming cycle, such as design, development, debugging and deployment.
"Architectures and programs are closely intertwined and a new architecture necessitates a new programming paradigm," said Dr. Dharmendra S. Modha, a key member of the team who developed the project at IBM Research. "We are working to create a FORTRAN for synaptic computing chips. While complementing today's computers, this will bring forth a fundamentally new technological capability in terms of programming and applying emerging learning systems."
IBM said its eventual goal is to construct a low-power chip system with 10 billion neurons and 100 trillion synapses, with a volume of less than 2 liters. This could lead to breakthroughs such as computer-assisted human vision via eyeglasses for the visually impaired.
The system is described in the following video:
A New Software Ecosystem for Cognitive Systems
David Ramel is the editor of Visual Studio Magazine.