SETI@Home Breaks 500,000 years

BoogieGod writes "The SETI@Home project has finally broken 500,000 years of computing time. They haven't detected any Extra Terestrials yet but there have been some interesting close calls. Now if only all 2.6 million of their users could join distributed.net."

Comments

Do something worthwhile, folding@home

[joshv]

500000 years? what a monumental waste. Why waste your CPU cycles on pie in the sky alien searches or breaking some encryption key you already know can be broken?

Simulate protein folding with your spare CPU cycles. It's a good cause, knowledge of how proteins fold helps determine the root cause of some genetic disease and can help researchers design better drugs.

Folding@Home

Granted, their screen saver kinda sucks, and there is no way to run the client without the screen saver, but I like the fact that I am contributing to a worthwhile cause.

Re:Do something worthwhile, folding@home

[Mr. Mikey]

Protein folding is certainly an important problem, and worthy of CPU cycles.

That said, it doesn't give you the right to diss those who want to contribute their cycles for the sake of the search for extraterrestrial life. A confirmed reception would have profound implications, far more than figuring out how a protein folds. So far, we haven't... and that does not invalidate the search.

Another project to help: YETI@Home

[f5426]

Check it here : http://www.phobe.com/yeti/

Cheers,

--fred

Why SETI@home

[tag]

I switched from distributed.net to SETI@home, and here's why: I know a key can be broken with brute force; I wonder if there's life on other planets. It's the mystery that draws me to SETI.

Re:Why SETI@home

[javatips]

That's also one of the reason I still give CPU time to SETI.

Instead of trying to break encryption stuff by brute force (hey who need to prove you can do it in xx time by actually doing it. it's a simple math formula, it can be proved in a few minutes). If distributed.net proprose to break some algorithm with some new techniques that may requires a lot of CPU but that need to be prooved that's it's more effective than brute force, then I will be glad to donate CPU time to distributed.net.

At least in SETI@Home, there is some science going on. These people try to prove something and they develop great tools and analysis techniques in doing so.

Put some science into distributed.net, then you'll have more users.

ps: add some pretty screen saver too, so users not interested in science or encryption can enjoy it.

Re:Forbidden Link

[Black Parrot]

> It seems the link for the "close calls" is forbidden?!? Anyone care to repost what it says somewhere?

I can get to it intermittently.

Spoiler alert...

Spoiler alert...

Spoiler alert...

It basically says that they got some funny clusters of spikes in their data, and couldn't explain them... until they noticed that the spikes' dates correlated with downtime for the 'scope.

--

Re:500.000 years computing time?

[rde]

Exactly what do they mean with 500.000 years computing time?
You (and others who were also bitching/wondering) are missing the point. The reason for celebration isn't that a x amount of years were done; nor is arguing that it could have been done in x/100 hours on decent machines a worthwhile consideration.
The point is that 2.5 million people, between them, have been running their computers collectively for half a million years. Doesn't matter whether this was on 8088s or top-of-the-range z80s, the owners' computers were running, in total, for that time. This is, IMO, a phenomenal achievement.

Don't be discouraged

[Cujo]

SETI at home hasn't officially found anything yet. What they mean by that is that they haven't found something that repeatably looks like a signal.

This doesn't mean that we're alone in the universe, for four reasons:

1. They're only looking at a frequency band where we would expect to fnd a signal if someone were deliberately trying to contact us. If someone were sending out beacons in random directions, then the signal wouldn't be straightforwardly repeatable. So, they'd have to know we were here. RF signals have only been transmitted from Earth for 100 years or so, and the vast majority of the energy in the last 50 years. So, only nearby civilizations (distance <= 50 LY) could know about us and be sending these signals.
2. The Arecibo antenna is actually a volcanic caldera, and can only sees a certain band of the celestial sphere, so there could be nearby civilizations transmitting, and SETI @ Home would miss them.
3. You can make a case that any civilzation capable of contacting us would almost certainly be far more advanced than we are. Given the way communications bandwidth is gobbled up by our relatively primitive culture, they'd probably be using all sorts of sophisticated spread-spectrum technology and a wide part of the electromagnetic spectrum that would make it very tricky for us to intercept and recognize messages not intended for us, even if we were lucky with the geometry.
4. The SETI@Home people are real scientists, and they kow that they need rock-solid evidence if they're going to claim they found a signal. So, they are bending over backwards - as they should - to find alternative explanations when they do see something anomalous.

The best thing about SETI at home is that it shows that you can harness vast amounts of computing power for a good cause with modest cost. Folding @ Home will hopefully get comparable attention.

Do something more useful...

[grytpype]

Try running the folding@home client instead. That project produces actual, useful scientific results about protein folding. SETI is just an inefficient search through a million billion haystacks for a needle that probably is not there.

Wisconsin , Cheese, and Idle Clock Cycles. . .

[Salgak1]

. . .would be the late Senator William Proxmire.

Who, with Walter Mondale, later to be VP, killed the "Station" part of the "Shuttle/Station" project in the early 1970's. That's right, we could have had a real space station, 20+ years ago.

And as for Idle CPU's. . .as I recall, Proxmire was a Democrat. . .

No OS/2 client either

[SpiceWare]

so I can't help them out either.

Checking out Seti's results by OS shows some interesting info. In the 90 OSes listed, MacOS is #3 in results and OS/2 is #20. Linux is #6. Based on SETI's results, F@H should have done the Mac client BEFORE the linux client.

shaekespeare@home:)

[bockman]

Given the assioma that a monkey, typing randomly, has a non-null chance to type in the Amlet (or your preferred novel of your preferred author), I propose a distributed computing project to do just that (after all, world's whole computer set should reach the same intelligence of a monkey)

what a waste of energy

[jilles]

There's been quite a lot of talk lately about PC and computer equipment being responsible for energy shortages. Seen in this context, the IMHO pointless and so far unsuccessful search for aliens seems a waste of energy. Computers are so much more energy efficient when suspended or turned off!

Re:what a waste of energy

[Junks Jerzey]

Agreed. If you're not using your PC for a while, just turn it off. With PCs sporting 450W power supplies these days, containing as many as five fans, and having video cards that get hot to the touch, turning the whole thing off is a good idea.

yeah, whatever

[Primer 55]

The important thing is that HALF A MILLION YEARS have collectively been spent processing the data. It is outstanding for the quantity of contribution that has been given to the project, not the quality.

Every thousand years
This little sphere
Ten times the size of Jupiter
Floats just a few yards past the Earth
You climb on your roof
And take a swipe at it
With a single feather
Hit it once every thousand years
Till you've worn it down
To the size of a pea
Well, I say that's a long time

The more interesting statistic... TeraFLOPs/sec

[sdo1]

This page indicates that in the last 24 hrs, 2.255873e+18 Floating Point Operations were performed on this project, or 26.11 TeraFLOPs/sec.

Wow.

-S

Arecibo (facts about)

[pq]

The Arecibo antenna is actually a volcanic caldera, and can only sees a certain band of the celestial sphere

I hate to nitpick, but Arecibo is not a volcanic caldera, in spite of what the tabloid press might report. In fact, it is a large limestone sinkhole in the karst terrain of Puerto Rico: check out this link for more info. (I promise its not a goatse.cx link.)

One of the cuter stories is that when they were searching for the perfect site on Puerto Rico, they took a dime and slid it around on a contour map of the island - and where it fit nicely inside the contours, there the dish went... Its amazing to look at, and I recommend a visit if you vacation in PR.

OTOH, your other point is completely correct - Arecibo only sees a limited range of the sky, and cannot view anything south of a certain declination (14? I forget). Not being able to see the Gal;actic center is particularly galling! That's why the new GBT (100m, unlike 305m at Arecibo, but the GBT is fully steerable) is so exciting.

distributed.net has a better project than RC5

[robertito]

Try OGR, the Optimal Golomb Ruler project. Finding better OGRs is actually a lot more clueful than brute-force cracking encryption keys (we've demonstrated that can be done, enough already!) - these interesting mathematical objects actually have many practical applications in comms, radio astronomy (so you are helping to find the LGMs) and other funky areas. And they even make beautiful necklaces....

Re:1 thing distributed.net lacks...

[cpfeifer]

The funny thing is the if you use the pretty seti@home sreensaver, you get severly degraded performance. With the screensaver it took my PIII-550 about 25 hours to chew through a block (running continuously). Just running the command line client (no pretty nothing), it chewed through 5-6 blocks per day. I easily blew away the competition and completed more blocks than 99% of the competition.

I dropped SETI@Home when I found out that they were recycling old blocks. What a waste of computing resources! My distributed.net client never runs out of good work to do.

How about a useful project (Climatic Modelling)?

[Malc]

I participate in Seti@Home, it seems more valuable than distributed.net's offering. However, I can see much better uses for such projects. A while ago I heard of the Casino-21 project, which is about climatic modelling. I think that this is something very useful, valuable and important to us, and I wish they would get a move on and start up. The web site of note for this project is: http://www.climate-dynamics.rl.ac.uk/. At this time you can only register your interest (I think that they're still in the planning phase.) I do recommend that people register as higher numbers might increase the chances of the project happening. Presumably registering will also sign you up for future announcements, such as when they have client software to download. Although they make comparisons to SETI@Home, I think they will operate slightly differently, with work units perhaps taking more than a year (one of the things that I think makes SETI@Home successful is that people get feedback via wu completion, and get to compete).

Extra Terrestrials?

[Black Parrot]

> They haven't detected any Extra Terrestrials yet

We've got some spares, if that's what you're looking for.

--

Meaningless statistic

[patreides]

Measuring computing time in years doesn't mean anything; these "years" are mostly being done on slow, outdated machines like at my school, where the average time is about 26 hours (that includes some G4 cubes, Blue G3's, but mostly really old all-in-ones from 5 years ago). Some of the machines like Sun's team does one every five hours or so along with SGI, but most of the time is from us slowpokes, it doesn't track any real data analysis or progress.

Re:What other services/researches are there?

[Tsuran]

There is a project to try and find a vaccine for AIDS (http://www.fightaidsathome.org) that works by running a specialized version of Autodock on your computer to simulate various molecules interacting with other molecules or what have you (I'm not a genetic engineer, the page is helpful).

I go for this one because it seems the most useful and important one out there. At the moment, it's Windows-only, but it shouldn't be too hard to find a Windows computer somewhere that the app can reside on. :)

Re:Because it's interesting to see

[Rader]

Although it wastes a lot of cycles, the screen saver has 2 very important rolex. One is to get as many people to do it. One thing we've all seen from this project: It's power in numbers. Whether you use the screen saver or not doesn't make much of a difference. But having maybe 400,000 of the people join up because they thought it "looked cool", makes a BIG difference. Even if they run the screen saver and waste cycles.

The other important role of the SETI screen saver is how it catches OTHER co-worker's attention. *MY* numbers might not be important, but the fact I got 10 other co-workers into it, and eventually became fanatics is.

I think a compromise of setting the screen saver portion to blank out after an hour is a good solution. No one (at least here) is going to see it running at night for 9 hours. Even if SETI@HOME was to rewrite it so it was twice as fast... it would only help them by a factor of 2, not a huge factor compared to the 2.6 million people.

Rader

Why bother with distributed.net?

[HEbGb]

I took a look at the current projects running from distributed.net, and couldn't find anything that appeared even marginally useful or interesting. Running brute-force hacks on encryption algorithms isn't much different than running a random number generator until your target value happens to appear. Both are equally useless.

At least SETI has a clear goal, and is a useful (and entertaining) pursuit which is naturally parallelizable. Other systems (PopularPower, etc.) also have useful things you can do with 'spare' cycles (at least if you're not the one paying the electric bill).

I fail to understand why anyone is advocating spending cycles on hunting for random numbers, a la distributed.net . Care to enlighten me?