This page was last updated on November 14, 2002

What is SETI?

SETI (Search for Extraterrestrial Intelligence) is a scientific project of the Berkeley University. The largest radio telescope on Earth, the 305 m Arecibo Telescope, is collecting data from parts of the sky that can be observed with this instrument. The amount of data is enormous and to process this data, it would require many supercomputers running continuously for many years.

The idea of SETI@home is to help process this data. Any person on Earth with a computer of practically any kind or speed, may process SETI data using a specially designed program which does that. More information about this is here and here.

Since no signs of intelligent life has been discovered so far, the only way to measure the involvement of each country, organization, or a person is to count how many data sets have been processed by this country, organization, or a person.

Puerto Rico (and our Department) joined the project the first day it officially started - on May 17, 1999. From the initial position number 46 among the 226 countries taking part in the experiment, we have slowly drifted to place number 50 in September 2001, and to 52 in February 2002. There are several countries with population smaller than Puerto Rico that are higher on that list: (New Zealand - 21; Ireland - 25, Lithuania - 42, Latvia - 46, Slovenia - 31, Estonia - 34, Luxembourg(!) - 47, and Iceland - 45).
In our Department 12 PC's in the Computer Lab (initially we had 15), plus two in my office, plus one at home, are running SETI. Our statistics is as follows:

September 4, 2001: the total number of users all over the world (each user may have more than one PC) is 3,247,171. Our position among these users is 1,842 (and we have a good chance to get a better position, since we are going up a few places every day). That means that 3,245,329 users are behind us.
September 8: number of users was 3,255,837. Our position: 1,831.
September 10: number of users was 3,258,240. Our position: 1,825.
September 18: number of users was 3,272,168. Our position: 1,786.
September 20: number of users was 3,273,840. Our position: 1,782.
September 23: number of users was 3,279,359. Our position: 1,776.
October 3: number of users was 3,292,993. Our position: 1,742.
October 8: number of users was 3,303,181. Our position: 1,717.
November 4: number of users was 3,350,786. Our position: 1,646.
November 11: number of users was 3,363,147. Our position: 1,636.
December 2: number of users was 3,405,735. Our position: 1,583.
January 3: number of users was 3,463,835. Our position: 1,505.
February 18: number of users was 3,550,198. Our position: 1,431.
February 25: number of users was 3,561,746. Our position: 1,420.
March 24: number of users was 3,615,828. Our position: 1,390.
May 6: number of users was 3,706,877. Our position: 1,370.
August 23: number of users was 3,928,359. Our position: 1390.
September 9: number of users was 3,959,219. Our position: 1405.
September 16: number of users was 3,971,838. Our position: 1399.
November 12: number of users was 4,083,142. Our position: 1359.
November 14: number of users was 4,086,434. Our position: 1350.

 
We have position number 1 in Puerto Rico at this time (we have moved from #3 to #2 on August 23, 2002, and to #1 on September 12, 2002).

Our position in the world is slipping from time to time - our highest position was 1,361 on August 10, 2002. Since then we started to move down slowly almost to 1400, now (14 November 2002) we are again higher, at 1350. The simplest explanation is that many new (and much faster) computers have been connected about the time when universities resumed work after summer break.
To see other statistics follow this link.

How you can help

Anybody with a computer may download the program from this place. Thre are versions specific to the operating systems, and even for different types of processors, so it is possible to run this program on practically every platform. There are also versions that display graphically what the program is doing, and versions that do not display anything but run about 20% faster.

How is the program running?

All versions of SETI@home run in the background with the lowest priority (idle). That means that this program runs ONLY if nothing else is running at the moment, starts automatically when the computer is doing nothing, but also stops automatically when ANY program starts. It is using then the time your PC is not doing anything important. Is the SETI program processor-intensive? You bet - but ONLY when nothing else is running. The question is - what is a running program? For example, when you surf the Internet, is the browser running? The answer is: yes and no. Yes - at the moment you open your browser, it is running for a few SECONDS, until it fully opens and displays the page. Then it practically stops, taking every few seconds only about 0.05% of the processor power, which means that practically it is not running. When you switch to another page on the Internet, your browser takes less than 10% of the processor power until the page is displayed fully, then it practically stops again (I have checked that!). Every program that you run on the PC behaves like that, except scientific programs processing large amounts of data and/or performing long and intensive calculations. Obviously games and simulators are in this category as well. These programs take over the whole power of the processor, but then SETI practically stops, as you can see in the next section.

How is SETI affecting operation of the PC?

It is a legitimate concern that such a program like SETI, running on your PC may slow down the computer and make all operations very slow. Because SETI runs with very low priority and stops each time another program starts, there should be practically no difference if SETI is running or not. If possible, it is always good to check that and express this with some kind of numbers.

I have performed tests on seven PCs, each having a different speed and a different processor. I have designed a computationally-intensive program and ran this program on each PC several times in two series: once with SETI running, and once with SETI removed from the system and not running.

The testing program was performing the following operations:
- a data file containing 4186 data points was read from the hard disk;
- from this data a segment containing 600 data points was cut out;
- a standard smoothing routine was used to smooth that segment of data;
- at each step a graph was displayed

Graphically, the testing program looks like that:

Results of the tests

The following table shows the results of running the above test on eight different platforms. Please note, that the testing program was running continuously - once with SETI running, and once without. The first four PC's were running Windows 98 Second Edition and Mathcad 2000 Professional. Two next PC's were running Windows XP Professional, and Mathcad 2000 Professional. The last two PC's were running Windows 2000 Professional and Mathcad 2001 Professional.

The following screenshot shows all running system components and programs and the CPU time that each of these processes is taking. The test program (MATHCAD) is running. The SETI program is NOT running. Please note the CPU time of MATHCAD. It is also interesting to note the CPU time that the mouse pointer is taking (POINT32.EXE). The screenshot was taken from a PC running Windows 98 Second Edition and Mathcad 2000 Professional. The processor was AMD K-6 III, 450 MHz.

The next screenshot shows the same thing, but this time the testing program is running together with SETI. Please note what is the CPU usage now for the MATHCAD and SETI programs. Note also the percentage that the mouse pointer (POINT32.EXE) is taking. The screenshot was taken from a PC running Windows 98 Second Edition and Mathcad 2000 Professional. The processor was again AMD K-6 III, 450 MHz.

Conclusion is obvious - running SETI does not affect normal operation of a computer more than running other programs. Actually, it's the other way around - running anything, even moving the mouse pointer, slows down SETI, because ANY program runs with a higher priority than SETI. A few percent of degradation can be noticed ONLY in case of CONTINUOUSLY running programs, which does not normally happen too often. I'm using three PC's - 350 MHz, 450 MHz, and 1.1 GHz and SETI is running even when I'm processing my pulsar data from Arecibo. I don't see any negative influence of SETI, even though some of my programs processing pulsar data are really computationally intensive.

It is possible that the test presented here is not objective and does not reflect the real situation. I would appreciate any suggestion to conduct a better test that would show that the results presented here are incorrect or reflect only a special case. If you have such a suggestion, please send me an e-mail, I would really appreciate that.