Many distributed computing
projects have taken the title of '@home'. This is in no way related to @home
, the cable modem network
but rather an indication that this project can be done at home.
One of the problems with some of the simulations and data analysis projects being undertaken today is the shear volume of data and decades if not centuries of computing time required to come up with an answer. Often, these projects are under-funded - SETI could not afford to purchase the time it has gotten on a super computer.
- The most well known '@home' project is that of SETI@home which as of Apr 4 has accumulated 616880.832 years of compute time (and is growing at 1100 years of compute time a day). This is on the order of 22.79 TeraFLOPs/second. The goal of SETI@home is to search the data from the Arecibo radio telescope
for signs of intelligent life. Most notably, it looks for strong signals that show up and fade away as the dish moves to point at other points in the sky, pairs and triplets of pulses.
(almost every platform in existence)
- Intel philanthropic peer to peer program
- Intel is also hosting a similar program that runs on Windows systems that looks at how drugs interact with four key proteins related to cancer and most especially the growth of leukemia. Working with United Devices the goal expected to be found at about 24 million hours of compute time and is likely to be the largest single computation chemistry project to date. Other projects hosted by United Devices include digitizing old information and genetic research.
(Microsoft Windows only at this time)
- Given a gene, it is easy to identify the amino acids that make up the protein. However, this is not enough - the very shape of the protein is important in how it interacts with with its target This is determined by protein folding. Implications of this work include looking at how HIV inserts its own genetic code
into cells, nanotechnology, understanding mad cow disease, and Alzheimer's disease. The project is only a few months old.
(Microsoft Windows, Linux, and Solaris currently, Macintosh
and other Unix flavors are being developed)
- Given how various rules about the physical and biochemical nature of how proteins behave, Genome@home strives to create new proteins (and thus new genes) that are not found in nature. By comparing these artifical genes to the real ones, researches can come to a better understanding of how the various proteins work and how the genes evolved. This has several real world applications:
- engineering new proteins for medical therapy
- designing new pharmaceuticals
- assigning functions to the dozens of new genes being sequenced
- understanding protein evolution
(Windows and Linux (x86 - requires GLIBC_2.2) At this time, a client for Macintosh OS X is being developed)
- A form of genetic computing that differs from most other distributed computing projects in that it creates its own data. Most other projects have a set of data that is analyzed - radio data from space, genetic code, etc... Here, the only limit is the 'imagination' of the random number generator. Here, the goal is to evolve a robot that moves forward. All of the walkers and rovers that have been thought of being sent have been based on designs that we know about - evolved over millions of years. However, this is only part of the immense possibility of 'how to move'. Part of the Golem@home project is to design a mover, and systems for designing things. Ideally instead of sending a rover to another planet, you would send spare parts, a computer, and a miniature factory that would design the rover for the environment that it is in.
(Microsoft Windows only)