Win a copy of Event Streams in Action this week in the Java in General forum!
  • Post Reply Bookmark Topic Watch Topic
  • New Topic
programming forums Java Mobile Certification Databases Caching Books Engineering Micro Controllers OS Languages Paradigms IDEs Build Tools Frameworks Application Servers Open Source This Site Careers Other all forums
this forum made possible by our volunteer staff, including ...
Marshals:
  • Campbell Ritchie
  • Devaka Cooray
  • Liutauras Vilda
  • Jeanne Boyarsky
  • Bear Bibeault
Sheriffs:
  • Paul Clapham
  • Knute Snortum
  • Rob Spoor
Saloon Keepers:
  • Tim Moores
  • Ron McLeod
  • Piet Souris
  • Stephan van Hulst
  • Carey Brown
Bartenders:
  • Tim Holloway
  • Frits Walraven
  • Ganesh Patekar

Limits on computing speed

 
Java Cowboy
Posts: 16084
88
Android Scala IntelliJ IDE Spring Java
  • Likes 1
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
In this video, professor Phil Moriarty explains what the physical limits are to the speed of computers, and the connection between computations, energy and loss of information (non-reversability). Especially that last thing seems to be what this forum is about.

 
Marshal
Posts: 65085
247
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Fred (and Bill, my first Director of Studies) will like the analogy drawn between computing and maths; they believe computing is an extension of maths, but I tend to think it is an extension of engineering: we are designing things. Who is correct Probably both.
There are several limits to speed of computation. One is that each loss of a bit releases kT•log(2) energy, which is to all intents and purposes 3×10⁻²¹J at about 40.5°C. That might be a small amount per bit, but it scales up pretty rapidly with faster computing; Pop says that the surface of a computer chip dissipates more heat per cm² than an electric hotplate on a cooker. Ten years ago, people like de Vos estimated that this irreducible heat production represents about 10⁻⁴ of the power consumpton of a chip; that has probably increased (as a proportion) by 10× since them. There is likely to be a limit to increasing performance as long as we continue to use silicon because at conductor sizes < 10nm (approx) tunnelling will occur and current will “hop” from one conductor to another producing incorrect results, but (Pop writes about this) new technolgies may make it possible to create much smaller conductors running at lower currents and consuming less power. But

each loss of a bit releases kT•log(2) energy

and this is unchangeable. Now your deletion of information consumes 10% of the chip's power, or even more, and that source of energy is a significant proportion of its heating costs (and its cooling costs). At present, cooling is the limit to computer performance, but this might change in the future.
There is much more to say, but not when people insist on wearing pink like that. At least he didn't have to confess it was his girlfriend who left it behind.
 
Campbell Ritchie
Marshal
Posts: 65085
247
  • Mark post as helpful
  • send pies
  • Quote
  • Report post to moderator
Just as well Seth Lloyd uses reversible gates to construct his ultimate laptop; running at 4.5×10⁵⁰ flop/s it would otherwise dissipate something in the region of 10³²W simply because of information deletion, and would become too hot to handle pretty quickly. Lloyd reminds us that there are other thermodynamic limits to computing speed and energy consumption: one is that it is necessary to expend energy in order to cause computation to proceed in a “forward” direction at a reasonable speed.
Lloyd later calculates heat dissipation because of error correction as 4.04×10²⁶W.
Phil Moriarty did say Lloyd's computer was a bit hard to handle.
 
Don't get me started about those stupid light bulbs.
  • Post Reply Bookmark Topic Watch Topic
  • New Topic
Boost this thread!