Jump to ContentJump to Main Navigation
The Age of EmWork, Love, and Life when Robots Rule the Earth$
Users without a subscription are not able to see the full content.

Robin Hanson

Print publication date: 2016

Print ISBN-13: 9780198754626

Published to Oxford Scholarship Online: November 2020

DOI: 10.1093/oso/9780198754626.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 17 October 2021

Scales

Scales

Chapter:
6 (p.79) Scales
Source:
The Age of Em
Author(s):

Robin Hanson

Publisher:
Oxford University Press
DOI:10.1093/oso/9780198754626.003.0012

Can we say anything about the specific speeds at which ems can run? Because of brain parallelism, the cost of running an em should be nearly proportional to speed over a wide range of speeds. The upper limit of this proportional-cost em speed range is the “top cheap” speed, that is, the highest speed at which the cost is still nearly proportional to speed. To estimate this speed, we must consider how simulated neurons in em brains might both send faster signals, and more quickly compute what signals to send. Human brain neuron fibers send signals at speeds ranging from 0.5 to 120 meters per second. In contrast, signal speeds in electronic circuit boards today are typically about half the speed of light. If signals in em brains move at electronics speeds, that would be between one million and 300 million times faster than neuron signals. If signal delays are the limiting factor in em brain speed, then this ratio gives an estimate of the maximum speedup possible, at least if em brains have the same spatial size as human brains. proportionally larger speedups are possible if em brains can be made proportionally smaller. Regarding the computation of when to fire a simulated neuron, note that real neurons usually seem to take at least 20 milliseconds to react ( Tovee 1994 ), while even today electronic circuits can switch 10 billion times faster, in one-and-a-half trillionths of a second ( deal et al. 2010 ). A key question is thus: how many electronic circuit cycles does it take to execute a parallel computer program that emulates the firing of a single neuron? For example, if there were an algorithm that could compute a neuron firing in 10 000 of these fastest-known circuit cycles, then an emulation based on this algorithm would run a million times faster than the human brain. As quite complex parallel computer programs can be run in 10 000 cycles, em speedups of at least one million times seem feasible, provided that energy and cooling are cheap enough to profitably allow the use of these fastest electronic circuits. When energy and cooling are more strongly limiting factors, however, the top cheap speed could be slower.

Keywords:   accidental, bacteria, communication, energy, frequency made, gravity, hardware, opaque, parallel computing

Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter.

Please, subscribe or login to access full text content.

If you think you should have access to this title, please contact your librarian.

To troubleshoot, please check our FAQs , and if you can't find the answer there, please contact us .