New Brain-Inspired Chip Can Perform 46 BILLION Synaptic Operations Per Second

New Brain-Inspired Chip Can Perform 46 BILLION Synaptic Operations Per Second

August 19, 2015 | by Aamna Mohdin
photo credit: Chip board. IBM
IBM researchers have been working on building a chip since 2008 that works like the neurons inside your brain. And they’ve just announced an exciting breakthrough. Scientists have developed a system that is made up of 48 million artificial nerve cells, which is about what you’d find in the brain of a small rodent.
The team has been working with DARPA’s Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) for several years now. They showcased the significant progress they’ve made with their TrueNorth system during a three-week educational boot camp for researchers and government officials. According to Wired, the TrueNorth system is a network of chips that has 48 million artificial nerve cells, with each chip containing 1 million artificial cells each. These chips are “neuromorphic,” which means they’re designed to behave like organic brains.
IBM researchers suggest that traditional computers work like the left side of our brain, similar to a fast number-crunching calculator. They compare TrueNorth to right side of our brain, likening the system to “slow, sensory, pattern recognizing machines.”
IBM researchers note that they “have not built the brain, or any brain” but have built “a computer that is inspired by the brain.” The TrueNorth system has been developed to run deep-learning algorithms, which is similar to the AI technology used for Facebook’s facial recognition or Skype’s instant translate mode.
The key difference is that IBM’s chips are a lot smaller, use less electricity and are cheaper to run. The TrueNorth system can therefore insert this AI technology into a much smaller package, such as a phone or wristwatch. TrueNorth’s 5.4-billion transistor chip uses 70 milliwatts of power, Wired reports. In comparison, a standard Intel processor with 1.4 billion transistors uses about 35 to 140 watts.  
“What does a neuro-synaptic architecture give us? It lets us do things like image classification at a very, very low power consumption,” Brian Van Essen, a computer scientist at the Lawrence Livermore National Laboratory, told Wired. “It lets us tackle new problems in new environments.”
It will take several more years before the chip will be available on the market, but according to IBM its unique architecture could solve “a wide class of problems from vision, audition, and multi-sensory fusion, and has the potential to revolutionize the computer industry by integrating brain-like capability into devices where computation is constrained by power and speed.”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s