Slashdot Mirror
SETI@Home to Crunch More Data
BigDave writes: "In this article on Wired, it describes how SETI is gradually running out of data, as the current data acquisition system cannot keep up with the rate of processing (since they now have 3 million users processing data). They have acquired a new high-speed digital data recorder which is Linux-powered, and was donated by Hewlett-Packard."
If there are not enough celestial data for the SETI@Home project, then let's turn some of that enormous Beowolfian processing power over to a categorically related AI@Home sub-project in the form of the First AI at http://sourceforge.net/projects/mind -- whjere we are creating the artificial intelligence that we may need (or may encounter) in the Search for Extra-Terrestrial Intelligence (SETI).
Just as the otherwise idle computers crunch data in the search for ET intelligence, the AI@Home project may become a contest to see whose computer will have the longest-running, gradually most ancient AI running as an uninterrupted artificial life (alife) form since Star-Date 200X.
A few hard-core AI@Homers may provide the algorithmic advances while the masses of participating SETI+AI enthusiasts provide the PC's, workstations and supercomputers.
When the AI@Home technology is sufficiently mature, then we turn the AI entities loose on the quest for their starborne brethren and sistren.
Logic dictates: lim --> *** (The stars are the limit.)
Maybe the Seti@Home project should consider re-crunching old data. Versions 3+ perform a LOT more calculations than 2.x or 1.x versions of Seti@home. How about adding a new 3.x version, that will only calculate the uncalculated portion of old data in the existing system.
===> An eye for an eye makes everyone blind - MG
SETI@home collects its data from the world's largest radio telescope in Arecibo, Puerto Rico. We've been recording data at the Arecibo telescope since December 1998, and analysing that data since May 1999.
SETI@home is a very fortunate science program. It utilizes 70% of the Arecibo telescope time. The other 30% is time used for repair, maintenance, or radar observations (Arecibo's powerful radar transmitters create too much interference for SETI@home's sensitive receiver).
This is an extraordinary amount of telescope time! Most astronomers are lucky to get even a day a year on the telescope for their research. Since SETI@home doesn't need to point to any specific point in the sky, it just "goes along for the ride" while other astronomers use the giant antenna. If SETI@home could take data full time we would collect about 50 GB of data every day. It takes us about eight months to "cover" the Arecibo sky. This isn't 100% of the sky that is visible to the telescope since we don't control pointing, but it's close. SETi@Home's goal is to collect and analyze at least two years worth of data. This would allow us to cover the sky seen from Arecibo about three times.
I've done the SETI thing and the Distributed.net thing and both, IMHO, were not very pragmatic. Other distributed projects exist, like Folding@Home and my favorite Genome@Home. They need more computing power, so please visit and try them. The even have Linux console versions for x86 machines.
There's even a $100,000 prize for the first 10,000,000 digit prime number. I encourage others to consider this project -- RC5 is close to pointless now (RC5/56 proved limited encryption is of no value), and SETI@Home already have more cycles than they can use.
I seriously question the science of SETI@home. I left them after one of the first debacles where they kept sending out the same packet of data to most everyone.
genome and folding@home just seems so much more likely to be useful.
If you're an atheist (or even if you aren't) you're welcome to join our genome@home team, Wicked Old Atheists. We're currently placed #24 in the world.
Belief is the currency of delusion.
For those who doesn't know it they (SETI@home) recently reached the Zettaflop (10e+21 floating-point operations) mark which is a world record. The last 24 hours "they" (read the users) performed 6.104916e+18 flops which is about 70.66 Teraflops/sec. This can be compared to the Terascale Computing System that theoretically could reach a maximum of 6 teraflops per second *laugh*. SETI's total cpu-time lies around 750 000 years, _pretty cool_ eh?
2 reptiles beneath your current threshold.
And it many ways, at that.
Consider the fact that we've had radios for a hundred years now, and TVs for quite a while now. Add to that cell phones and satellite communications, and you've got a nice big EM bubble around Earth, of radius 100 light-years (since EM travels at light speed, and we've been sending them out for a century).
Granted, a hundred light-years isn't much, but if aliens within that distance are looking out for signals in the same way that we are, they've got quite a large source of incoming info.
But there's more! On March 15, 1999, a 400 000 bit-long transmission was sent out to four "local" star systems suspected of harbouring life. Take a look at the fascinating Encounter 2001 transmission. It's absolutely worth a look. Try to figure some of it out too, just for fun =) IMO, it's brilliant.
So we are, after all, broadcasting quite a lot, whether it be specific targetting or general.
Cheers.
BTW S@H have admited for a long time that they send out each unit 3 or 4 times, for double-checking, and because they aren't splitting/recieving the units from Arecibo fast enought. However they only use a small band of Arecibo's datastream, centered on the H-OH 'waterhole' (1420MHz +/- 1.25 MHz); this should improve the rang of frequencies covered.
There is talk of using southern SERENDIP as a second antenna to get better sky coverage. They have another problem; S@H accounts for about 30% of Berkeley Uni's total out going bandwith, outside the Space Science Lab, the net admins aren't that happy about this. Unless they can get other SpaceScience Universities to share the load, they can't increase their userbase much more.
It has become appallingly obvious that our technology has exceeded our humanity. --- Albert Einstein
All newer transmissions systems, from SSB to spread-spectrum to GPS to HDTV, don't use carriers. The FCC wouldn't license a transmission system today that used a carrier. In time, all radio will be carrierless, to save spectrum space. That date is probably about 20 years away, after the transition to HDTV and digital audio broadcasting. So for less than a century will our civilization have broadcast carriers. That's a narrow window to hit when looing for another civilization.
There's some redundancy in all carrierless systems, but it may be only a few percent, and it's hard to find if you don't know how to look for it. Typically, detecting a spread-spectrum signal involves trying to synchronize a psuedorandom number generator at the receiving end with the signal. This is hard when you have no idea what the psuedorandom number generator looks like. It's not impossible; it's a cryptographic problem. But it's hard to detect a signal so weak you can't read the bits.
You can look for the presence of a carrier so weak that you can't detect the modulation, by averaging over many cycles. That's what SETI@Home actually does. So if there are carriers out there, SETI@Home should find them. But unless someone is deliberately beaming carriers at us, there's nothing to find.
I've met some of the SETI@Home people, and they admit this problem. By now, if anybody in our stellar neighborhood was aiming high-power continuous carriers at us, we'd know it. But there could be signals encoded in more efficient ways and thus look like noise. SETI@Home will never find them.
I think that the SETI@Home effort should be devoting more resources to finding non-carrier signals. Maybe long-period autocorrelation, looking for repeats of bit patterns, would be more appropriate than the present carrier search. Something that sounds like stellar hiss might turn out to have data in it.
Of course it's. I think many will disagree but there's no point breaking RC5 or any other cryptographic key. We already know that it's possible. It's like breaking a glass. The only question is if it will break with the first hit... or how long it will take to break the key - we already know it'll happen sooner or later anyway. If we weren't breaking the key with brute force it could be more interesting...
Searching for ET is more interesting because we don't know the answer for sure. Probably we won't find anything. OTOH, why miss the change to be the discover if we do?
In the end, helping with folding problem would probably be the sane thing for a geek because there's a nice probability that we get something usefull out of used CPU time.
_________________________
Spelling and grammar mistakes left as an exercise for the reader.
I prefer projects with a higher probability to make an actual differene to how people live, like the (already named) Folding@Home, Genome@Home, or FightAIDSatHome. The last one may not appeal to many here as Entropia, the distributed computing network behind it, apparantly insists in throwing in some commercial work packets to the clients. Finding a cure for AIDS sounds like a splendid idea, otoh.
My personal favorite is GIMPS, the Great Internet Mersenne Prime Search, discoverers of the four largest explicitly known prime numbers. I like them because you actually have a chance to understand what the program is doing (if number theory is for you, that is). IMHO better than looking at some blinking lights of a screen saver that looks for ET.
Alex
Heisenberg may have been here