GPU : A Gnutella Processing Unit

Interesting distributed computing projects

Three interesting (and professional J) distributed computing projects are presented here: they are called Seti@home, The Cancer Research Project and Unlike GPU, they are based on a centralized model.

1. Fast fourier transform to find E.T.

Nowadays, there are huge amounts of data recorded from telescopes. The rough data needs some processing before scientists can evaluate it. In particular, the 305-meter radio telescope in Arecibo, Puerto Rico searches the sky for special events in the Universe around the frequency of 1420 MHz. This is the spectral frequency of hydrogen, the most common element in the universe, and therefore a universal reference point for different intelligent species.

The researchers at the Space Sciences Laboratory of the University of California, Berkeley started the distributed computing project because the university was low in funds and could not buy another faster supercomputer. The SETI@home project (SETI = Search for Extraterrestrial Intelligence) is the most successful project in this field. In August 2002 there were four million users registered. They compute at a speed of about 40 Teraflop/s. The speed of the fastest supercomputer on Earth, the nipponic Earth Simulator is only about 36 Teraflop/s.

The algorithm used is a fast Fourier transform over chunks of data of about half a megabyte, a recording a one and a half minute of the sky. The fast Fourier transform tries to point out gauss curves with a basis of about 12 seconds and finds spikes and triplets in these data chunks. The scientists do not exactly know what to search for; they just can guess how alien signals might look.

Development of satellites moves towards the principle that data will be analyzed on ground (where a variety of different algorithms can be applied) and not on-board as before. This is possible thanks to the improvements in communication technology.

Article: Know what to expect when E.T. calls

Link to Seti@Home

The Arecibo radiotelescope in Puerto Rico. It’s sensitivity is about 10-13 W/m2 (photo NASA)

2. Research against cancer

Another important project we want to present in this document is called LIGANDFIT. The following text is quoted from the Internet site of The Cancer Research Project :

“The program you download allows your computer to screen molecules that may be developed into drugs to fight cancer. Each individual computer analyzes a few molecules and then sends the results back over the Internet for further research. This project is anticipated to be the largest computational chemistry project ever undertaken and represents a genuine hope to find a better way to fight cancer.

Through a process called "virtual screening", special analysis software will identify molecules that interact with these proteins, and will determine which of the molecular candidates has a high likelihood of being developed into a drug. The process is similar to finding the right key to open a special lock—by looking at millions upon millions of molecular keys.

Participants in the United Devices Cancer Research Project are sent a ligand library over the Internet. Their PC will analyze the molecules using a docking software program called LigandFit by Accelrys. The LigandFit software analyzes the molecular data by using a three-dimensional model to attempt to interact with a protein binding site. When a ligand docks successfully with a protein the resulting interaction is scored and the interactions that generate the highest scores are recorded and filed for further evaluation.”

Once again, a project over the peer to peer network could perform a similar task.

Link to the project site

3. Break strong cryptography is the last distributed computing project presented here: the project breaks cryptography standards using brute force attacks and sometimes wins competitions issued by security companies. The client also searches for Optimal Golomb Rulers.

Here's an explanation of what Golomb Rulers are:

Golomb rulers are nothing more than lineals with a small number of marks. Optimal rulers are the ones that with n marks can measure the most distinct distances, if they are used only by looking at the distance between two marks on the lineal. Note that to place one mark means to repeat it for the whole length of the lineal. If we want to search for optimal rulers the marks must be located very efficiently, so that we avoid redundant distances between marks. Additionally, perfect rulers are optimal rulers that can describe distances between 1 and n. A theorem described in an old Scientific American issue (“Mathematical Games”, where the work of Golomb was published first) states that perfect rulers can exist only for n<=6. The last perfect ruler has 4 marks.

Golomb rulers are used in radioastronomy to place a minimum number of antennas.

On August 2002 the search is open for optimal rulers with 24 and 25 marks.

Article: In search for Golomb rulers

Link to the project site
Seeschloss created the initial website. The logo was designed by Mark Grady. Graphics are by David A. Lucas.
GPU, a P2P-computing cluster software, © 2002-2015 by the GPU Development team