A chip that works like a human brain consumes 16 times less energy

Researchers at Graz University of Technology and Intel have shown that neuromorphic computing has

equipment has a huge potential to run complex deep neural networks DNN.

According to them, neuromorphic computing hardware can run deep neural networks 4 to 16 times more efficiently than conventional hardware.

During the experiment, the team evaluatedenergy efficiency of a large neural network that ran on a neuromorphic computing chip from Intel. This DNN was specifically designed to handle a large number of numeric and alphabetic sequences.

The researchers measured howthe power consumption of an Intel neuromorphic chip and a standard computer chip when running the same DNN, and then compared their performance. The researchers found that if the neuron models in the computer resembled the neurons in the brain, then this helps to improve energy efficiency.

In their next research, the authors want to givethe ability for neuromorphic equipment to develop a personal memory based on its past experiences. Just like a person does, using individual experience to make decisions.

Read more

The Japanese dumped a giant turbine into the ocean to get endless energy from the current.

Astronomers from Japan have found an unknown structure in the galaxy

Researchers filmed a 'hidden' ecosystem in an Antarctic river