• Subscribe

iSGTW Opinion - Volunteer computing: grid or not grid?

Opinion - Volunteer computing: grid or not grid?

BOINC has captured the imagination of hundreds of thousands of volunteers, many of whom donate not only their computing power, but also their time to BOINC projects. BOINC fans write poetry, design logos and compete to produce the best BOINC results.
Image courtesy of Jared Hatfield

Around the world, one million computers participate in volunteer computing, providing 1.5 petaflops of computing power to scientists.

Is this grid computing? David Anderson, director of SETI@home and founder of Berkeley Open Infrastructure for Network Computing (BOINC), says "no."

"Volunteer computing" allows people to donate the use of their PCs to science projects. Early examples include GIMPS, Distributed.net, SETI@home, and Folding@Home. Today, volunteer computing is being used in biology, climate study, epidemiology, physics, and more, and most of these projects use BOINC software.

"Grid computing" lets organizations such as companies, research labs, and universities share computational resources. Globus and Condor are widely used grid software.

Is volunteer computing a form of grid computing? Both are forms of distributed computing that try to more fully utilize existing resources. However, they differ in several essential respects:

Volunteer computing must embrace amateurs
Volunteered resources are owned and managed by regular people, not by IT professionals. Therefore the software must be simple to install, it cannot require specific operating system versions or configurations, and it must be self-healing.

The BOINC phenomenon crosses the borders of discipline, nation, and language. This poster comes from the Japanese BOINC application "TANPAKU," which uses volunteer computing time to study protein structure.
Image courtesy of TANPAKU

Volunteer computing must tolerate hackers.
Volunteers are anonymous, and those who misbehave-perhaps by returning incorrect computational results or claiming credit for work not actually done-cannot be fired or prosecuted. Software for volunteer computing must accommodate the possibility of such misbehavior by, for example, running each job on two computers and comparing the results.

Volunteer computing "pulls"; it does not "push"
Volunteered resources are primarily behind network firewalls that do not allow incoming network connections. This requires the use of a "pull" model in which PCs periodically request work from a central server, rather then the "push" model used by most grid software.

Volunteer computing is a one-way affair
Grid computing is generally symmetrical: an organization can borrow resources one day and supply them the next, in the same way that utility companies trade power on the electrical grid. Volunteer computing, in contrast, is asymmetric: volunteers supply computing resources to projects, and not the other way round.

Volunteer computing pays Internet Service Provider bills
Volunteer computing uses the "commodity Internet," where both projects and volunteers must pay for network bandwidth. This mandates careful planning for data-intensive applications.

Volunteer computing demands great public relations
Volunteer computing relies on volunteers. Thus scientists can access volunteer computing power not by requesting or purchasing allocations, but by persuading the public that their research is worthwhile. This public outreach is a significant fringe benefit of volunteer computing.

Thus, although journalists will no doubt continue to confuse them, volunteer computing and grid computing are very different beasts. They have different capabilities, use different software, and require different skills and mindsets from their users.

- David P. Anderson

David P. Anderson is a research scientist at the University of California, Berkeley.

Join the conversation

Do you have story ideas or something to contribute? Let us know!

Copyright © 2019 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.


We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.