Michael Stiber, Ph.D.

Member since March 10, 2023

Business Type: UW Faculty

Business Name/Affiliation:
UW Bothell

My research focuses on developing an understanding of biological computation by examining simulations of nervous systems. I am currently working with students to develop large-scale neural simulations using GPU (Graphics Processing Unit) hardware, which harnesses the number crunching capability behind video games and animated movies for general-purpose computation. This is being applied to models of neural development, formation of network connectivity, and the subsequent behaviors that are produced.

In an additional project, I have been working with a colleague at the University of Montreal at Quebec on what we’ve termed “Sapien Systems”. The motivation is a simple observation of the growth in computer performance over time. Between 1990 and 2010, the cost of one megabyte of disk storage went from $9 to $0.00015. Had the human population followed a similar growth curve, there would be 300 trillion people on Earth. Obviously, growth in computer performance will continue to far outstrip human population growth. One could conclude from this that any algorithm that is related in size to human population should not be evaluated in terms of asymptotic complexity — that in fact it is constant-time. What kinds of problems are like this (for example, perhaps all problems relating to the web)? Which classes of problems are larger? Assuming that most real-world problems are those of sapien systems, what are the implications of computing capability that renders them constant-time?

Listings by Michael Stiber, Ph.D.

Research & Creative Projects

BrainGrid+Workbench: Bringing High Performance Neural Simulation to the Desktop

Added on September 8, 2023

Listings by Michael Stiber, Ph.D.

BrainGrid+Workbench: Bringing High Performance Neural Simulation to the Desktop

Sign In

Register

Reset Password