Intel scales up self-learning neuromorphic computing

Intel scales up self-learning neuromorphic computing

Technology News |
By Peter Clarke

Intel launched Loihi in 2017 as a 14nm FinFET chip capable of representing 130,000 neurons and 130 million synapses (see Intel launches self-learning processor). Unlike convolutional neural network (CNN) and other deep learning processors the Loihi test chip uses an asynchronous spiking model to mimic neuron and synapse behavior in a much closer analog to animal brain behaviour.

The Pohoiki Beach system comprises one or more Nahuku boards, each of which contains 8 to 32 Loihi neuromorphic chips.

The Pohoiki Beach is capable of representing 8 million neurons and is suitable for research project on such things as sparse coding, graph search and constraint-satisfaction problems. “Pohoiki Beach will now be available to more than 60 ecosystem partners, who will use this specialized system to solve complex, compute-intensive problems,” said Rich Uhlig, managing director of Intel Labs.

Chris Eliasmith, co-CEO of Applied Brain Research and professor at University of Waterloo, said that Loihi had been able to demonstrate a 100 times lower power consumption running a real-time deep-learning benchmark, compared to a GPU and 5 times lower power than specialized IoT hardware.

Later this year Intel plans to introduce a larger Loihi-based system called Pohoiki Springs capable of representing 100 million neurons.

Related links and articles:

News articles:

Intel launches self-learning processor

French startup develops self-learning AI chipset

Consortium seeks to scale artificial intelligence

BrainChip launches neuromorphic hardware accelerator

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles