It's not a widely known fact but according to the Guinness Book of Records, the record for the largest computation project in history isn't held by either a government or corporation. The record goes to David Anderson's seti@home volunteer project, hosted by the Space Sciences Laboratory, at the University of California, Berkeley. So far, the distributed processing application seti@home has racked up over two million years of aggregate computer time, nearly all of that volunteered by individuals at home and at work.
Even if seti@home has yet to discover extra-terrestrial intelligence, BOINC has already discovered something useful to the advertising and marketing industry. Until seti@home came along, there was no way for a SMB to sponsor any significant research and development. Seti@home, and its next generation version BOINC, has made low to no cost R&D sponsorship a viable option for any size business. Not surprisingly, this discovery was not by accident.
In 1999, seti@home was launched with two goals; to prove the viability and practicality of the 'distributed grid computing' concept, and to do useful scientific work by supporting an observational analysis to detect intelligent life outside Earth.
The first goal was critical because seti@home had an enormous challenge to overcome in its search for extraterrestrial life. While seti@home was able to get large quantities of data from the Arecibo radio astronomy telescope in Puerto Rico, the bottleneck lay in processing it. Every 15.5 hours the telescope generated 33 gigabytes of data, and sifting through it looking for the frequency spikes and other telltale signs of extraterrestrial life took vastly more computer processing power than seti@home could afford.
Seti@home's solution was highly unique and a totally game changing innovation. Seti@home designed an Internet (aka grid) based computer system that would allow people to volunteer processing time on their personal computer. SETI designed a small screensaver for people to download and run on their computer which would automatically log on to the SETI@home server, download a .35MB block of data (about 107 seconds worth), process the data and return the results to the SETI server.
To reward participants, SETI set up a system to track to and publish the hours of work done by each participant, and promote competition between them. One of those encouraged by the prospect of low cost favorable publicity was Microsoft. They promptly created their own version of the seti@home
application, optimized to run on Windows, and thus prove the superiority of Window systems by generating and recording more work in less time. After careful analysis, SETI decided the Microsoft modifications to their software were actually cheats, rather than real improvements, and disqualified recognition for work done with the Microsoft version of seti@home
. This was the first known case of a company trying to cook the books to get undeserved recognition for contributing to a grid R&D project, but it won't be the last.
Even without Microsoft's help, the success of seti@home and BOINC drew the attention of the scientific community and in 2006 a new and much more powerful and versatile grid system replaced the original seti@home
. The new system was born at Berkeley as the Berkeley Open Infrastructure for Network Computing, or BOINC for short. The BOINC system offers a grid system similar to seti@home
, but will support many different projects through a single screensaver application. Volunteers simply download the screensaver and then select whatever combination of projects they wish to donate time for, and the BOINC system does the rest.
The BOINC system has proven popular with scientists, in part because it substantially reduced the cost and complexity entering the market of grid computing. Currently, there are over 30 different projects available through BOINC, including the fields of biology, medicine, astronomy, physics, chemistry, climate modeling, earth sciences and mathematics.
The project has also proven popular with the public. BOINC had hoped to get between 50,000 and 100,000 volunteered computers, but by 2009 had over 2.4 million computers available with over 270,000 active at any time. However, the number of hours volunteered by people has been declining recently, in part due to ever increasingly restrictive IT department policies at larger businesses. This is would be an excellent time for SMBs to step in to take up the slack.
The primary cost to a business for supporting BOINC is the cost of the extra electricity used. The BOINC application can be installed, and uninstalled, in a few minutes. As with any network based software, compatibility problems sometimes show up, but they can be minimized if handled properly.
Implementing a grid R&D sponsorship project for your business is a very straightforward process.
- Step one. Download BOINC from boinc.berkeley.edu and try it on one of your computers for a few days to a week. If there are no problems, install it on most, but not all, of your other machines.
- Step two. Determine which projects you want to donate processing time to, and how much time will go to each. Remember that while the BOINC screensaver will run on nearly any Mac, Windows or Linux system, the actual projects themselves are frequently limited as to what platforms they will run on.
- Step three. Create and formally document your company's policy for using BOINC. Just state what hours the BOINC system can and cannot be operational, and exactly what projects time will be allocated to.
- Step four. Set up a process to routinely monitor and compare the performance and operational cost of the systems running BOINC to the machines without it (look for a spike in the power bill, or in network down time). Be ready to disable BOINC if it conflicts with other needs of your business.
- Step five. Promote your efforts. Put a banner ad on your Website's homepage and in your company's promotional material that announces the fact your business is donating processing time in pursuit of knowledge. Promote the different projects you are supporting, and if feasible, list the number of hours you have donated to each so far.
Don't miss this one. The evolution of grid based R&D sponsorship is significant for two reasons. It offers a new way for SMBs to generate goodwill with the public, and it marks the first time major R&D projects were executed independent of budgetary control, or the influence of special interests, from either major business or governmental institutions. Neither may be as earth shaking as finding extraterrestrial intelligence, but it the long term they may be far more useful to us.
Glen Emerson Morris was recently a senior QA Consultant for SAP working on a new product to help automate compliance with the Sarbanes-Oxley law, an attempt to make large corporations at least somewhat accountable to stockholders and the law.
He has worked as a technology consultant for Yahoo!, Ariba, WebMD, Inktomi, Adobe, Apple and Radius.