Analogy of Distributed Computing

Message boards : The Lounge : Analogy of Distributed Computing
Message board moderation

To post messages, you must log in.

AuthorMessage
JordanWeber

Send message
Joined: 24 Apr 08
Posts: 1
United States
Message 16836 - Posted: 24 Apr 2008, 18:53:23 UTC

Does anyone know if there is an (obviously incorrect, but close to) analogy of how much you are contributing by running your computer for a day?

Like: say for instance, "running your computer with <insert bio/folding project here> for 1 day is equal to finding a cure for <a certain or well known disease> 1 second faster"

"Thus, if only 1000 people participated, in 1 year, we could find this cure 4.2 days faster, or 33.6 days faster with an "average" computer."

Obviously this is incorrect, but it helps to show what an impact a person could have for a project, and helps them convince their friends/family to get involved as well.

Some results should be stated like: "assumes 1 cpu at 500mhz with 75% idle usage" and it would take factors like redundancy checks.
ID: 16836 · Report as offensive
mo.v
Avatar

Send message
Joined: 13 Aug 06
Posts: 778
United Kingdom
Message 16846 - Posted: 25 Apr 2008, 2:13:44 UTC
Last modified: 25 Apr 2008, 2:21:06 UTC

Hi Jordan

It might be like that if progress in science happened at a steady rate, but it doesn't always. In some fields the researchers keep plodding on, analysing data and so on, and sometimes may find big things, other times small things, and sometimes nothing at all. Or they may spend time replicating experiments and either finding the same thing again, or sometimes invalidating earlier work. And occasionally someone hits on a new concept that rapidly changes the very way people think. I'm thinking of the chaos theory developed by Edward Lorenz who died recently.

In a few cases your approach might discourage people from crunching. Say 5000 people use their 5000 computers to crunch 5000 climate models that on average take 4 months running 24/7 to complete. A group of scientists uses the ensemble to produce a publication of 5000 words. Each cruncher might think they had contributed one word. A person whose slow computer battled with the data, crashed it several times, restored it, then completed it after 7 months might not be very pleased to think that 'their' word could turn out to be 'the' or 'a'.

But if they learn something about the climate and numerical modelling while running the task, read the abstract of the publication and perhaps see the research or the topic referred to in newspaper articles or on television, they'll realise that it's important and probably be glad they had the opportunity to participate.

Participating is of course usually more fun if it's easy for beginners, so we need to do whatever we can to ensure that it is.
ID: 16846 · Report as offensive

Message boards : The Lounge : Analogy of Distributed Computing

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.