Brain Power

Scientists at IBM research unveil a brain-inspired computer and ecosystem

What is cognitive computing?

Cognitive computing aims to emulate the human brain’s abilities for perception, action and cognition. The neurosynaptic chip, designed to emulate the neurons and synapses in the human brain, breaks path with traditional architectures used for the last 70 years.

Click the plus signs for additional information, including photos and video.

+

Introducing a Brain-inspired Computer

TrueNorth's neurons to revolutionize system architecture

Dharmendra Modha

By Dharmendra S. Modha

Six years ago, IBM and our university partners embarked on a quest—to build a brain-inspired machine—that at the time appeared impossible. Today, in an article published in Science, we deliver on the DARPA SyNAPSE metric of a one million neuron brain-inspired processor. The chip consumes merely 70 milliwatts, and is capable of 46 billion synaptic operations per second, per watt–literally a synaptic supercomputer in your palm.

Along the way—progressing through Phase 0, Phase 1, Phase 2, and Phase 3—we have journeyed from neuroscience to supercomputing, to a new computer architecture, to a new programming language, to algorithms, applications, and now to a new chip—TrueNorth.

Traditional computers focus on language and analytical thinking (Left brain). Neurosynaptic chips address the senses and pattern recognition (Right brain). Over the coming years, IBM scientists hope to meld the two capabilities together to create a holistic computing inteligence.

Unprecedented scale

This second generation chip is the culmination of almost a decade of research and development, and is a huge leap forward from the initial single-core hardware prototype developed in 2011.


Programmable neurons
2011 to 2014. 256 to 1 million.

Programmable synapses
262,144 to 256 million.

Neurosynaptic cores

1/10th of a Watt powers the neurosynaptic chip's 256 million synapses
...with the goal to simulate 1 trillion synapses using only 4kW of energy

Different from a standard chip

Traditional chips run all of the time. This new neurosynaptic chip is event-driven and operates only when it needs to, resulting in a cooler operating environment and lower energy use.

The neurosynaptic chip veers from the traditional von Neumann architecture, which inherently creates a bottleneck limiting performance of the system.

+

The chip is especially designed for low power consumption, which can clearly be seen in this thermal image that shows the cool TrueNorth chip alongside hot FPGA chips that are feeding data to TrueNorth.


New architecture

IBM's brain-inspired architecture consists of a network of neurosynaptic cores. Cores are distributed and operated in parallel. Core operate -without a clock- in an event-driven fashion. Cores integrate memory, computation, and communication. Individual cores can fail and yet, like the brain, the architecture can still function. Cores on the same chip communicate with one another via an on-chip event-driven network. Chips communicate via an inter-chip interface leading to seamless scalability like the cortex, enabling creation of scalable neuromorphic systems.

+

Traditional computers separate memory from computation, requiring bits to continually shuffle between memory and the CPU via a bus. This means power must increase as the communication rate (clock frequency) increases. In contrast, the brain operates a low “clock” frequency and with low power density.

In IBM’s new TrueNorth chip, computation, memory, and communication are intimately integrated, which leads to low power operation.


Ecosystem

IBM has developed an end-to-end ecosystem for developing applications on these brain-inspired chips that includes a simulator, a programming language, sample algorithms/applications, a library, and a teaching curriculum.

+

The chip is one element of an entire software-hardware ecosystem for creating scalable neurosynaptic computing systems.


Where does this lead us?

IBM's long-term goal is to build a neuro-synaptic chip system with ten billion neurons and one hundred trillion synapses, all while consuming only one kilowatt of power and occupying less than two liters of volume.

10 billion neurons. 100 trillion synapses.

This technology will be used in many fields that span both research and industry, including public safety, vision assistance for the blind, home health monitoring and transportation.

+