Technology

The World's First Photonic Neural Network Is Really, Really Fast

Virtually every device you use—from the one you're using to read this to the pocket calculator growing dust in the back of your desk—relies on the same basic technology: circuits containing many tiny transistors that communicate with each other using electrons. We've come a very long way since the room-sized computers of the 1950s, but as computing gets smaller, faster, and more complicated, we get closer to hitting a wall. There's a physical limit to how powerful traditional computers can get. That's why scientists are turning to completely new forms of technology for future computers.

For the first two parts of our three-part series on the future of computing, we explored what computers might be like if they thought like humans or used photons to move information. Now find out what happens when those two technologies become one.

The Best Of Both Worlds

Neural networks make for computers that can think and learn. Using photons instead of electrons in a computer makes data travel at the speed of light. Put them together, and you get one seriously powerful computer. That's what Princeton University researchers unveiled when they published a Cornell University Library report in November of 2016: the world's first photonic neural network.

How It Works

The network relies on a silicon chip, just like a conventional computer, but instead of transistors, it features 49 circular nodes that act like neurons. Each of these "neurons" circulates a specific wavelength of light, then releases it. That released light has an effect on a laser, which completes the circuit when its light returns to the neuron.

As complicated as that sounds, this should read clear as day: "The researchers proved that the chip is capable of super-fast computing by demonstrating that it could crunch a mathematical differential equation 1,960 times more quickly than a typical central processing unit, which uses electrons," says Futurism. That's right: by swapping out conventional transistors for artificial neurons and comparatively sluggish electrons for light-speed photons, scientists created a computer that's nearly 2,000 times faster than the ones we use today. The future is looking bright—and really, really fast.

Is there something you're curious about? Send us a note or email us at editors (at) curiosity.com. And follow Curiosity on Facebook, Instagram and Twitter.

Watch And Learn: Our Favorite Content About The Future Of Computers

How Close Are We To Computers That Think Like Humans?

A new breakthrough may have gotten us closer.

All About Quantum Computing

The sky's the limit.

Share the knowledge!

How A Neural Network Could Change Everything

Once computers can learn, who knows what's possible?

Written By Curiosity Staff March 3, 2017