Finally, from the information conveyed we will draw parallels between the Brain and our present-day computers, which will expose the reasons why our current approaches are falling short of doing what the Brain does, and will conclude with a call-to-arms for the computing, nano, and VLSI communities."> Finally, from the information conveyed we will draw parallels between the Brain and our present-day computers, which will expose the reasons why our current approaches are falling short of doing what the Brain does, and will conclude with a call-to-arms for the computing, nano, and VLSI communities."> TUD, chair Cuniberti "materials science and nanotechnology" - Lehrstuhl Cuniberti "Materialwissenschaft und Nanotechnik"
Skip to content.

TUD

search  |  internal  |  deutsch
Personal tools
TU Dresden » Faculty of Mechanical Science and Engineering » Institute for Materials Science » Chair of Materials Science and Nanotechnology



Thursday , 11 July 2013
(at 13:00 in room HAL 115)
Add to your Google Calendar


Why Biology Can ... and ... Silicon Can't? The brain and the computer

Valeriu Beiu


United Arab Emirates University
   






This presentation aims to follow on the steps of von Neumann's prescient The Computer and the Brain (Yale University Press, 1958), and to clarify common misconceptions about the Brain. The tale will be reversed, i.e., we shall start from the Brain, and rely heavily on the latest results, for trying to foresee the future of bio-inspired computing. We will shed some light on the elementary nano-devices of the Brain, and also on how we could decipher its intricate architectures for information processing. These will require covering quite some ground, parts of which are still being explored and not well understood, hence implicitly charting directions for research -- both for the computing as well as for the VLSI/nanotechnology communities.
We will start from the gated ion channels, and go on to argue about the different ways they "communicate." Revealing the statistical (collective) behaviors of arrays of gated ion channels for communications will follow. For computations, we shall make the case for interweaving arrays of various gated ion channels. In support of these we will detail preliminary and kind of unexpected results on the extremely high reliability of such multiarrays. Touching upon the ultra-low (but reasonably constant) power consumption of the Brain, we will stress the crucial role the ion pumps are playing here. Synapses (the gaps between neurons) and glial cells (e.g., astrocytes) will be briefly mentioned, before delving into the higher levels of the Brain, trying to grasp their optimal hierarchical organizations -- emphasizing columnar structures as large fan-in cyclic circuits (i.e., with feedback).
Finally, from the information conveyed we will draw parallels between the Brain and our present-day computers, which will expose the reasons why our current approaches are falling short of doing what the Brain does, and will conclude with a call-to-arms for the computing, nano, and VLSI communities.

Brief Bio:

Dr. Valeriu Beiu received the MSc in CE from the University "Politehnica" Bucharest, and the PhD summa cum laude in EE from the Katholieke Universiteit Leuven. He was with the Research Institute for Computer Techniques, University "Politehnica" Bucharest, Katholieke Universiteit Leuven, King's College London, Los Alamos National Lab, Rose Research, Washington State University, and since 2005 he is with the United Arab Emirates University. He was a PI/co-PI of over 30 research grants totaling over US$ 30M, holds 11 patents, has authored over 220 papers, and has given over 170 invited talks. His research interests are on VLSI/nano-electronics (low-power, highly reliable) and biological-/brain-inspired nano-architectures (parallel, adaptive, fault-tolerant, communication starved). Dr. Beiu received five fellowships, including Fulbright, EU Human Capital and Mobility, LANL Director's, and Rose Research. He was an Associate Editor of the IEEE Transactions on Neural Networks and is an Associate Editor of the IEEE Transactions for VLSI.

Announcement (pdf)



slides (pdf)

Invited by G. Cuniberti

Within the nanoSeminar

last modified: 2018.10.24 Mi
author: webadmin