Energy efficiency and the design of brains

Location: CPD Room, M413, Merz Court
Date/Time: 3rd May 2012, 11:00

Speaker: Simon Laughlin. See: http://www.neuroscience.cam.ac.uk/directory/profile.php?SL104

An understanding of design, the efficient organization of components into purposeful systems, helps neuroscientists to reverse engineer brains. Systems that sense, recognise, predict and control have to consume energy to process information, and brains are no exception. Energy budgets show that energy consumption limits cerebral signal traffic, indicating that there is powerful pressure to evolve energy efficient designs. I will review and present a number of design principles. These principles account for some of the distinctive features of neurons, neural circuits and codes, such as sparse coding in parallel pathways, the ability to regulate structure and allocate resources according to demand, and computing with chemistry. Perhaps these will help us to understand how a 1.5 kg human brain can computer prodigiously with < 20W of power?

published on: 23rd April 2012