Distributed computing

Tricord · Post by **Tricord** » Thu Jun 23, 2005 4:52 am

I was wondering, as I was reading a thread about distributed computing.. Has any project paid off anything yet? I mean, with the sheer amount of teraflops made available by the public, these projects are bound to have some results (except Seti@home, of course. That was ridiculous).

How about folding@home and other worthy causes? Have they yielded any results or have they just been sucking up CPU power and achieve nothing?

I'd love to hear from people who follow these things closely.

roid · Post by **roid** » Thu Jun 23, 2005 5:28 am

wow mate, i was just thinking the exact same thing, probabaly reading teh exact same thread.

i was looking at that gravity-waves distributed computing thing and noticed that it all seemed to be arranged like a COMPETITION? wtf. all i want to do is help, to be a part of it*, i don't wanna do it faster than anyone else - 2 ppl racing to do the same computational unit just wastes resourses.

* so my prime motivation is watching it move forward, so a finish line would be nice. ANYTHING goal based.

Genghis · Post by **Genghis** » Thu Jun 23, 2005 6:26 am

The GIMPS project's objective is to find Mersenne primes, and has found about half a dozen Mersenne primes so far. Definitely positive results.

http://www.mersenne.org/freesoft.htm

By the way, the GIMPS project is almost a decade old! The only other distributed computing project I know of that old is SETI.

Mobius · Post by **Mobius** » Thu Jun 23, 2005 2:49 pm

SETI at home isn't a joke mate: they are deadly serious. SETI is now human-kind's largest ever computational project - more CPU cycles have been devoted to SETI than any other project, ever.

I've got about 9000 WU's on SETI now - and I'll keep running it too.

The chances of finding any signals is non-zero, and any non-zero number is worth taking a punt on. (Look at LOTTO tickets buyers!)

Personally, I'm runnign SETI just to eliminate the possibility that any other civilisation is beaming something at us, or radiating such a huge amount of power that we'llpick it up.

I believe the chances are absolutely tiny however, as us looking for radio waves is like Sioux Indians trying to intercept Telegraph signals by looking for smoke on the horizon, or by pressing their ear up against the copper cables carrying the signals.

No self respecting civilisation would bother to use the elctromagnetic spectrum to commmunicate - so anyone we *DO* pickup, will likely (have been!) in a similar technological state as ourselves.

Top Gun · Post by **Top Gun** » Thu Jun 23, 2005 6:28 pm

So, Mobius, what other spectra of phenomena would you suggest that another civilization use? Unless you've suddenly advanced physics by a greater degree than Einstein himself did, I don't know of any other feasible way outside the realm of pure science fiction.

I would personally view Folding@Home to be a much more worthy use of computational time than SETI@Home, or even the Einstein or math-related programs. While I'm all for pure scientific discovery, the possibility that protein folding could aid in finding a cure for some of humanity's most deadly diseases is more than enough reason for me to single that effort out.

roid · Post by **roid** » Thu Jun 23, 2005 7:17 pm

Top Gun wrote:...I don't know...

bingo.

"top gun doesn't know" is a lot different to "it will never happen".

Zoop! · Post by **Zoop!** » Thu Jun 23, 2005 7:51 pm

Mobius wrote:The chances of finding any signals is non-zero, and any non-zero number is worth taking a punt on. (Look at LOTTO tickets buyers!)

Mobius, trust me. I sell lottery tickets at my summer job and let me tell you, you do not want to associate yourself with lotto ticket freaks. You have not seen a crybaby who thinks the world is going to collapse in on itself until you tell one of those people that the lottery machine is down. I'm not saying that all lottery ticket people are evil. I'm just saying that Darwin needs to get back to work.

Other than that, these projects do look like fun. Most people's computers are on all the time anyways...

De Rigueur · Post by **De Rigueur** » Thu Jun 23, 2005 10:16 pm

roid wrote:
i was looking at that gravity-waves distributed computing thing and noticed that it all seemed to be arranged like a COMPETITION? wtf. all i want to do is help, to be a part of it*, i don't wanna do it faster than anyone else - 2 ppl racing to do the same computational unit just wastes resourses.
.

What's wrong with competition? It can motivate people to do better and it makes the work more interesting while you're waiting for a positive result.

BTW, if two people were doing the same computational unit, it would be a waste whether they were racing or not.

Jagori · Post by **Jagori** » Thu Jun 23, 2005 11:47 pm

Well they need to have some way to make sure that the work unit was computed properly, no? It'd be a little sketchy to base medical research on unverified data.

Tricord · Post by **Tricord** » Fri Jun 24, 2005 6:43 am

What the hell is SETI computing anyway? We know what folding does: it does molecular simulation. The premises are simple; establish rules and give input data, then simulate. This may take time when accurate results are required, much like ray-tracing an environment.

SETI doesn't simulate anything, it merely "analyzes" data. I for one have no idea how a computer can spend hours and hours "analyzing" a unit of data that took a few seconds to download.

I have always thought of SETI as the biggest hoax out there. What do we hope to find anyway? With our current measurment capability (satellite, telescope, etc..) I would think that automatic analysis and processing is still much faster than the rate at which the data is generated by our sources. Distributed computing is not needed at all.
Furthermore, you still need a human to interpret things.

On a sidenote, Mobi you should relativise. You make too many assumptions and then you get carried away by some crackpot sci-fi story.
I've said it before and I will say it again: the extraterrestrial life problem can be solved with statistics. And I'm not even interested in the results. ET stay home!

roid · Post by **roid** » Fri Jun 24, 2005 9:31 am

ah double checked. ok

Lothar · Post by **Lothar** » Fri Jun 24, 2005 11:17 am

Tricord wrote:What the hell is SETI computing anyway?

Lots and lots of Fourier transforms.

You don't understand how it could take a long time to do a bunch of Fourier analysis on a small (~300K) signal? Maybe you should sue your school, then.

Tricord · Post by **Tricord** » Fri Jun 24, 2005 11:26 am

Well, let me rephrase my idea including math babble, let's assume we have a certain signal decomposition given by a certain Fourier transform. What the hell are we looking for? The ground problem stays the same, before or after all Fourier transforms you want.

PS: We don't sue at the drop of a hat over here.

De Rigueur · Post by **De Rigueur** » Fri Jun 24, 2005 2:21 pm

Tricord wrote: What the hell are we looking for? The ground problem stays the same, before or after all Fourier transforms you want.

This "ground problem" -- is it the problem of distinguishing signal from noise? . . . the problem of recognizing something as being intelligently designed as oppose to being the random by-product of natural causes?

If so, I think you have a point. I mean, if we're going to say that something as complex and ordered as, say, human DNA is not the product of intelligent design, then even if we received a signal with similar order and complexity, how would we know if it was just random or evidence of intelligence?

Tricord wrote: PS: We don't sue at the drop of a hat over here.

Hey, tell that to Microsoft.

Top Gun · Post by **Top Gun** » Fri Jun 24, 2005 7:24 pm

roid wrote:
Top Gun wrote:...I don't know...
bingo.

"top gun doesn't know" is a lot different to "it will never happen".

True, but I would guess that most modern physicists would share my opinion. Last time I checked, no one has figured out how a subspace transmission is supposed to work.

Lothar · Post by **Lothar** » Sat Jun 25, 2005 1:30 pm

Tricord wrote:Well, let me rephrase my idea including math babble, let's assume we have a certain signal decomposition given by a certain Fourier transform. What the hell are we looking for?

Interesting patterns of Fourier coefficients. In this case, "interesting" means "extremely unusual"...

Kyouryuu · Post by **Kyouryuu** » Sat Jun 25, 2005 3:46 pm

I used to participate in SETI@Home, but when there are folding protein projects and networked programs investigating cancer, it just seems like a waste of CPU time. I have profound doubts that if SETI ever found anything, they'd announce it. They'd merely face the wrath of the government lynch mob.

It's not that I don't believe in extraterrestrial life. But we've got some problems on our own blue marble with cancer and disease. Those are tangible issues with big ramifications. Yes, space is fascinating. But until we've got things under control down here, we're ill-suited as a species to go out there.

Sirius · Post by **Sirius** » Sun Jun 26, 2005 12:01 am

Top Gun wrote:
roid wrote:
Top Gun wrote:...I don't know...
bingo.

"top gun doesn't know" is a lot different to "it will never happen".
True, but I would guess that most modern physicists would share my opinion. Last time I checked, no one has figured out how a subspace transmission is supposed to work.

True, and perhaps they never will. But we don't know that. Simply because humanity doesn't know of it doesn't mean it doesn't exist, or scientific advancement would be impossible; we'd already know everything.

roid · Post by **roid** » Sun Jun 26, 2005 6:25 am

topgun wrote:True, but I would guess that most modern physicists would share my opinion. Last time I checked, no one has figured out how a subspace transmission is supposed to work.

yes, how many of those modern physicists would say "it'll never happen" topgun?
few to none.

almost everything we know now would have been considered "rediculous" back at one time in scientific history. so scientists who know their history know to never say never.

last time you checked no-one has figured out how a subspace transmission was supposed to work.
turn the clock back a few hundred years topgun, and no-one had figured out how an internal combustion engine is supposed to work. yet here they are.

how are your subspace transmissions any different? think about it

Genghis · Post by **Genghis** » Sun Jun 26, 2005 8:59 am

I thought gravity waves had potential for non-EM communications.

Ned · Post by **Ned** » Sun Jun 26, 2005 1:02 pm

Revolutionary Grassroots Astrophysics Project "Einstein@Home" Goes Live
http://einsteinathome.org/
http://www.physorg.com/news3103.html

Good conversation! My 2 cents:

communications: maybe laser, maybe radio, maybe something we have never discovered. It all seems plausable.

I didnt know there was a folding project. Hopefully it helps important proteins and the results are public, not some drug company's.

Since we humans haven't figured out better ways to do math, some of the stuff is very clumsy. Fouriers, drawing boxes under integrals, etc., but it's not a hoax. For example LIGO/Gravity/Einstein:
http://einsteinathome.org/ask/archive/index.html

Even though the detectors are stationary, it uses doppler corrections and fourier stuff to look "point" to regions of the sky.

I think the SCORING thing is just a way to make people excited and be able to track their contribution. Einstein@home has a graph of your data quantity analyzed. It's fun to see it increase.

Heck, maybe its a good way to see if your overclocking is done right. Mine runs 100% CPU time, and my machine seems happy. (No overclocking though)

Tricord · Post by **Tricord** » Sun Jun 26, 2005 4:07 pm

Subspace? What the hell is subspace, other than a cheesy Star Trek term?

Top Gun · Post by **Top Gun** » Sun Jun 26, 2005 8:08 pm

Damn, I'm really getting taken to task here for attempting to make fun of Mobius.

I realize all of that, roid; I plan on becoming one of those physicists myself. It's just that Mobius's seeming certainty that humanity's use of the EM spectrum for communication marks us as a "primitive" species, and that a more "advanced" species would definitely use some method that science hasn't even begun to discover, was somewhat amusing to me. I have to start watching my mouth around here.

Tricord, as far as I know, that's about all it is right now, but as an avid fan of Star Trek and FreeSpace, I'm hoping that there's something more to it than just fiction.

roid · Post by **roid** » Sun Jun 26, 2005 10:05 pm

well we're already getting past the EM spectrum.

remember all that talk about quantum entanglement (i think it was entanglement, tricord would know. or maybe you'd know since physics is of interest to you) being used for communications? where 2 particles were being linked through some quantum physics principle, then could be taken far apart from one another yet would STILL influence one another's properties.

the theory was to use such principles for EMless wireless communications.

Tricord · Post by **Tricord** » Mon Jun 27, 2005 3:46 am

Yes, quantum entanglement.. But there is a catch. Suppose you have a pair of entangled photons that can be in two states, but it is not yet determined which. If you keep one foton to yourself and give the other to your partner, observing your own foton will determine the state of both. However, you can't choose which state, you just know both states as soon as you know one. As of yet, there is no way to actually impose one state over another in order to yield useful communication. But it is nonetheless some form of communication, yes.

Anyway, I think I can summarize this thread as follows: SETI and similar projects have NO results at all, folding has minor results but no real breakthroughs, some prime research projects have found a handful of new primes... That's about it. Given the CPU time donated worldwide to these projects, those are VERY feeble results in my opinion...

roid · Post by **roid** » Mon Jun 27, 2005 4:46 am

someone should design some peered distruted computing virus or somesuch, try to evolve a self-aware digital super-organism.

suicide eddie · Post by **suicide eddie** » Mon Jun 27, 2005 6:34 am

Roid said "someone should design some peered distruted computing virus or somesuch, try to evolve a self-aware digital super-organism?"

explains a few peeps on the dbb

Top Gun · Post by **Top Gun** » Mon Jun 27, 2005 1:09 pm

I've heard about quantum entanglement to a small degree; it's a very interesting phenomenon. I'm not sure if it can be practically applied to any sort of large-scale communications application; from what I understand, it's a rather new line of research, and as Tricord said, there are catches involved. Still, maybe something can come out of it in the future.

Ned · Post by **Ned** » Mon Jun 27, 2005 1:30 pm

Tricord wrote:t know both states as soon as you know one. As of yet, there is no way to actually impose one state over another in order to yield useful communication. But it is nonetheless some form of communication, yes.

Im not sure thats true:
http://www.newscientist.com/channel/fun ... rld/dn5076

if you use rapidly flip flopping polarization detectors, I think you can encode "information" into it, like bits. You can measure in a pulsed way, which can overlay info . . . I think. Im no expert

Sirius · Post by **Sirius** » Mon Jun 27, 2005 5:34 pm

Top Gun wrote:It's just that Mobius's seeming certainty that humanity's use of the EM spectrum for communication marks us as a "primitive" species, and that a more "advanced" species would definitely use some method that science hasn't even begun to discover, was somewhat amusing to me.

Well, I do get your point. Mobius tends to assume a lot of things, although most of the time they are fairly safe even if they don't sound like it.

roid · Post by **roid** » Mon Jun 27, 2005 9:50 pm

Ned wrote:
Tricord wrote:t know both states as soon as you know one. As of yet, there is no way to actually impose one state over another in order to yield useful communication. But it is nonetheless some form of communication, yes.
Im not sure thats true:
http://www.newscientist.com/channel/fun ... rld/dn5076

if you use rapidly flip flopping polarization detectors, I think you can encode "information" into it, like bits. You can measure in a pulsed way, which can overlay info . . . I think. Im no expert

not quite what i was talking about, but close.
that article talks about quantum states of photons traveling down a physical fibre optic pipe.
what i was talking about was NO PIPE. is it called quantum tunneling? where the actual mystical LINK between the 2 quantum entangled photons is utilised to transmit data. with no physical link (no wires)

Ned · Post by **Ned** » Mon Jun 27, 2005 10:28 pm

roid wrote:
not quite what i was talking about, but close.
that article talks about quantum states of photons traveling down a physical fibre optic pipe.
what i was talking about was NO PIPE. is it called quantum tunneling? where the actual mystical LINK between the 2 quantum entangled photons is utilised to transmit data. with no physical link (no wires)

I hear you. The confusing part is what if a photon smacks an oxygen molecule and changes course, does it's twin change course? That would seem to violate cause and effect. But I think a lot of the NO PIPE stuff is done in a very cold vacuum. Zelinger thinks that he can get molecules as big as C60 fullerene to be in TWO places at once! If the he's right, we are in for a weird ride Dorothy.

http://www.bnl.gov/bnlweb/pubaf/pr/2001/bnlpr020601.htm

""Anton Zeilinger (born on 20 May 1945 in Ried im Innkreis) is a (Someone who is a member of the faculty at a college or university) professor at the (Click link for more info and facts about University of Vienna) University of Vienna, previously Innsbruck. Zeilinger is known for multiple experiments in the realm of quantum interferomerty, the most famous being the demonstration of (Click link for more info and facts about quantum teleportation) quantum teleportation. His work influenced the (Click link for more info and facts about experimental) experimental progress in a new sub-field of physics - (Click link for more info and facts about quantum information theory) quantum information theory. Also, he is widely known for many works in theoretical physics and the foundations of quantum mechanics - the most influential one showed a dramatic amplification of the Einstein-Podolsky-Rosen paradox, where one considers three, not just two entangled particles.""

His website isnt loading for me, but his articles have great titles and stuff
http://www.ncbi.nlm.nih.gov/entrez/quer ... eilinger+A