“I’m an engineer by training, but a scientist at heart,” Stanford University’s Kwabena Boahen told the crowd at our DC headquarters last week during the final public program of our spring season.

Throughout his talk, “Neuromorphic Computing,” Boahen emphasized how critical it is to use engineering to solve scientific problems.  

And the obstacle he is trying to overcome is a doozy—something that will affect us all.

For the past 50 years, computers have been getting smaller and more-compact, Boahen explained.

“By shrinking these devices, we’re able to use less power and this is how now you have what in the 80s was a supercomputer in your pocket in your phone,” he said.

But engineers have reached the point where they can’t make transistors any smaller without drastically increasing their price. What’s more, Boahen added, about a decade ago engineers stopped decreasing the amount of voltage needed to move electrons through a transistor. 

“Even though we can put more transistors on a chip, we can’t turn them all on at the same time, because they will burn too much power and the chip will melt,” he said. “So, we are hitting all these different roadblocks and it’s actually the first time in 50 years that it has happened … quite dramatic.”

This means that for the first time in a half-century computer engineers are hitting a wall of fundamental physics, and new approach is necessary if society wants technology to continue to improve by the same leaps and bounds to which we are accustomed.

Enter Boahen’s field—creating computers that work like brains.

It sounds like something straight out of a futuristic movie where robots take over the world, but Boahen assured us that there is no need to worry.

“We are very far from getting to the point where we can build this sci-fi-type robot that’s intelligent, that’s moving around, and interacting with the real world and outsmarting us, because he’s going to need about 20,000 watts of power, he’s going to be weighed down by his batteries … and he will run out of juice in a few minutes,” he joked.

Boahen and other neuromorphic computing experts are studying the way that charged particles move through miniscule channels in our brains, which is more efficient than the way that electrons move through computer transistors.

“Biology has found a way to basically squeeze everything down to one single lane and be able to still work,” he said, something computer engineers would like to mimic one day. 

When asked how long it will be until we see some of the products of his research in our daily lives, Boahen said that we will need to deploy some of what he is studying if we want to make the next major technological leap forward.

It’s like we’re elite athletes who have been using a certain technique to swim, but have reached the fundamental limit of how fast the human body can go unassisted, he explained. Now we need to wear a pair of fins to set new speed records.  

Boahen believes his research will be those fins. We hope he’s right!