Sign In

Communications of the ACM

ACM TechNews

Thought Experiment: Build a Supercomputer Replica of the Human Brain


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
An image of brain activity.

The Human Brain Project eventually will need at least 100 petabytes of RAM and an exaflop of computing power to enable its simulations.

Credit: University College London

Henry Markram believes his Human Brain Project can simulate all 86 billion neurons in the human brain as well as the 100 trillion connections among them, an ambitious effort that will require unprecedented computing power.

Essentially Markram intends to develop a plug-and-play brain that researchers could disassemble to determine the causes of brain disease, apply to robotics to develop intelligent technologies, or combine with virtual reality glasses to experience another person's brain.

The project’s first Blue Gene supercomputer could simulate a single neocortical column in a rat, but an entire rat brain has the equivalent of 100,000 columns. The Human Brain Project will eventually need a minimum of 100 petabytes of RAM and an exaflop of computing power to enable its simulations.

Critics say Markram's project is doomed to fail and question the $1.3 billion in funding the European Commission awarded the project in January. However, supporters defend the Human Brain Project, believing that even if Markram does not accomplish all he aims to, the project will make tremendous advances in neuroscience.

"There aren’t any aspects of Henry’s vision I find problematic," says the University of Manchester's Steve Furber. "Except perhaps his ambition, which is at the same time both terrifying and necessary."

From Wired News
View Full Article

 

Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA


 

No entries found