"Nightlife" Harnesses Idle Fedora Nodes For Research

A. B. VerHausen writes "If you've given up on SETI, now you can let your idle computer help with other kinds of scientific research. Red Hat employee Bryan Che started a project called Nightlife. He wants people to 'donate idle capacity from their own computers to an open, general-purpose Fedora-run grid for processing socially beneficial work and scientific research that requires access to large amounts of computing power.'" Che hopes to have more than a million Fedora nodes running as part of this project.

SETI

[pryoplasm]

There is also folding at home http://folding.stanford.edu/ [stanford.edu] that might help someones life more than software ever will.

I am all for open source, but there are some better places to donate some spare cpu cycles

Re:SETI

[abolitiontheory]

Agreed here. I don't have much hope in the mission of SETI, but Folding@Home's research is basically like throwing a gigantic brute force attack at unsolved protein mysteries. It feels like hacking, in a way. I love that idea, instead of just processing bombarded information from outer space.

Re:SETI

[somersault]

unsolved protein mysteries

You mean like the chicken and the egg debate? Why pot noodles are so addictive? Or how that mouldy block of cheese managed to solve the game of solitaire that you'd left lying out on the kitchen table?

Re:SETI

[morgan_greywolf]

Not to mention climate change prediction at home via climateprediction.net [climateprediction.net].

Re:SETI

[mwvdlee]

I guess that's irony; letting your computer consume large quantities of environmentally unfriendly produced electricity in order to calculate climate change.

Re:SETI

[somersault]

It's better to think of it as job security.

Re:SETI

[deroby]

Errr, yes they are !

When Folding@Home runs in the background my cpu is 100% all of the time (well, one core is in each case). When it's not running, I average around 10% I guess.
The difference is that in the latter case the cpu runs pretty much idle for 90% of the time and needs some electricity to keep going, while the former situation has it working at full throttle all the time, consuming so much more energy that the generated heat needs to be actively removed from the portable. I'm not saying it draws 10 times the amount of power, but it's going to be considerably more !

All that said, I often wonder what would be more efficient : 10.000 specialized cpu's in some server-farm / data-center churning away on a given problem, being mostly limited to that single problem and costing heaps of money and energy, or 10.000.000 versatile grid-clients that more or less produce the same output, probably eating just as much energy, if not more.

Re:SETI

[TClevenger]

That's why I don't participate with the hardware I have. A PS3 is 10-20x the speed of my system in those calculations, while using about 1.5x the electricity (~200W vs. ~125W for my desktop.) Those with certain graphics cards can use a GPU client and also get 10-20x the performance, for only slightly more power consumption than my desktop.

Hopefully they get a multi-GPU client polished soon. Then I would consider buying two or three supported GPUs and running a megaclient, but only when I need the heat anyway (i.e. wintertime or cold nights.) Right now, running a desktop client in my office would be even worse--I would need A/C to remove the extra heat.

Re:SETI

[bcat24]

But for most people, they don't turn on their computer just for that anyway. They just let it run as screensaver when they're away for a while, or they let it to make use of unused CPU cycles while they're working or doing other stuff. It's not like they're using extra electricity

Except that they are. Modern CPUs have quite a bit of power-saving support built in. If the processor is idle for a long period of time, it will practically shut itself down. Don't get me wrong, I like the Folding@Home project, but it does cause a noticeable increase in your electric bill.

Re:SETI

[Krigl]

I'd personally still prefer Folding@Home - climatology is way too complex, with lots of unexplained and speculative stuff. I'm not a scientist, but I'd guess this needs more basic research of underlying principles before brute force number crunching starts yielding useful results (any climatologist here?), not mentioning this project screams "junk science" out loud. And if they want internet community to get interested maybe someone should enlighten them about possibilities of different picture formats than 22 MB .bmp for high resolution histogram of global temperature change [climateprediction.net].

Folding@Home is useful and brings actual results - you'll get a chance to throw your own pack of frozen pea against Africa's hunger, instead throwing it into wastebasket of "well, it seemed as a way to go then".

As for SETI, well, yes there's a lot of space research fans here and way more Star Trek and Star Wars fans, who just secretly wish aliens to exist because it would be so cool if they existed even if without a chance to get into a hot threesome with Spock and E.T, but let's face it - aliens don't exist. And if they do, hoping to get some proof from SETI is like going to the sea coast once in your life, step on the shore with closed eyes and reach into the water in hope you'll get a grasp of bottle with a message from boat wreck survivor.

If you gonna donate spare cycles, donate them on something useful instead of something cool or guilt relieving.

Re:SETI

[Otter]

Folding@Home is useful and brings actual results - you'll get a chance to throw your own pack of frozen pea against Africa's hunger, instead throwing it into wastebasket of "well, it seemed as a way to go then".

The gap between Folding@Home and anything that addresses "Africa's hunger" is at least as speculative as anything in climatology, and not that much less than SETI.

Re:SETI

[Otter]

As for Folding, it would be enough to get some better meds, even the current-like without side effects would be great result.

Again, getting to that from Folding@Home results is at least as speculative as climate modeling.

Re:SETI

[LingNoi]

Please explain to me why you think protein folding is as speculative as trying to predict the climate of an entire planet.

Re:SETI

[Otter]

Beyond the fact that, as J.Y. Kelly explains, protein folding is an insanely complex problem, the assertion was variously that Folding@Home results would directly lead to solutions for "Africa's hunger" and "some better meds", both of which are rather wild extrapolations from what the project actually produces.

Re:SETI

[phulegart]

Wow.. here I was, picking out posts that needed some moderation (since I'm sitting on points)... then I came across your post. So much for moderation.

I can't decide if I should just reword what you've got there, in the standard /. tradition, or if I should just deal with the issues I see one at a time.

Climatology is way too complex, with lots of unexplained and speculative stuff. Ok. I'll agree to that. You aren't a scientist? Ok. I'll tuck that fact away for future use. How would you have any clue as to what you should designate as "Junk Science" if you are not a scientist? As someone who is not a scientist, how would you have any clue as to how much more research needs to be done before ANY number crunching is done? Does it not occur to you that this number crunching you refer to *is* research? Shouldn't you be a scientist in this specific area before you condemn their actions in attempting to propel advancements in their field? As someone who admits NOT to be a specialist in climatology, what the hell do you know?

And if they want the internet community involved, they should what?? Get Enlightened as to different image formats? Why? Because you don't like the size of that bmp? Because you realize that they could pass out even more finely detailed images that would be smaller, if they only know what YOU know? Do you already know ALL of the reasons they chose to use BMP? Did YOU inform them of alternate image format possibilities that provide them with ALL the advantages they believe they are getting from BMP images, while also getting other advantages YOU are aware of that they are apparently not aware of? Have you been in contact with them at all, about this image format problem that you see? Or are you waiting for someone ELSE to read your mind as to the details of this issue, and waiting for someone ELSE to send off an email that climateprediction.net should change formats? I mean, I get it. You want someone other than yourself to make a fuss about this BMP thing. You can find the time to post about it here, but you can't be bothered to mention the problem to the people who matter.

Now... where is the address at the site Stanford put up to support Folding@Home? You know the address I'm talking about... the one that I can mail my frozen peas to, to feed starving folk in Africa. Because I didn't find it. No, you can't get away with saying that you were just pointing out how using Folding@home is LIKE actually sending food to end African hunger. I do get it that you are CLEARLY saying here that using anything other than Folding@Home (in your opinion) is a complete waste of time and energy... your wastebasket reference. But on the whole, and in parts, you are wrong. Folding@home is no more, or no less a waste of time than SETI@Home... or any other distributed processing that is being passed around. Remember, you are not a scientist. You have no idea if the numbers you are crunching are going toward a problem that is going to be solved in 5 years, or in 50 years... or at all. However, for all you know, data crunched for Climateprediction.net could provide positive and useful results in 2 years. You admit that you are not a scientist, or a climatologist. So you are simply an ignorant individual who is claiming to have all the right answers.

Now... SETI. Ok. Fine. You don't believe that there is intelligent life out there in the Cosmos. You have done a good job in proving to me that there was no intelligence at work behind the creation of your post. So, since I see no evidence of thought or intelligence in your post, I should assume there is no intelligent life on Earth? Show me your proof that we are the only intelligent life in the Galaxy... or the Universe? I won't leave you hanging... you CAN'T prove it, because it is IMPOSSIBLE to prove that we are the only intelligent life out here. However, it is the height of arrogance and hubris to assume that we are the MOST intelligent species out here in the universe. We are on the v

Re:SETI

[phulegart]

Granted you quoted and repeated me... although while I just said that it can't be proven that we are the only life out here, you did a nice job of pointing out that you can't prove a negative.

You did lose sight of facts and details in my post (probably because it was too long for you to read)... I can prove that aliens do exist. In fact, let me rephrase that... it can be proven that Aliens do exist. SETI is in the process of proving that. If you choose to see the fact that it has not been proven YET as concrete evidence that it will NEVER be done, you fail to remember your history. I don't even have to mention any one specific historical fact, because there are countless examples of where something had not been done, that same something was called impossible to do, and then it was done.

I'm not saying that Aliens exist. I am also saying that they probably do exist. However, you know Air exists although you can't see it. You know that microbial bacteria exist, even though you can't see, or touch them. You know that there are planets orbiting other stars, even though you personally can't see, touch, taste, smell, or hear them.

But which probability is higher... the probability that there is other sentient life, somewhere out there in the universe, or the probability that we are the ONLY sentient life in the entire universe?

Re:SETI

[bWareiWare.co.uk]

You realise folding at home is software right?

Re:SETI

[an.echte.trilingue]

I thought the most interesting part of this is that he thinks he can get a million people to do this.

Fedora is mostly a hobbyist OS (as opposed to RHEL), and I bet a lot of Fedora machines are desktops. If that number is at all realistic, the number of Linux users worldwide is way underestimated.

Re:SETI

[Danathar]

If you don't know what the guy is talking about, then don't comment.

Condor is WAY different than BOINC or Folding@home.

BOINC is middleware but NOT general purpose grid computing. Condor is a distributed batch oriented system that allows people to submitt jobs and get them done. You can configure BOINC to run as backfill to Condor when Condor is not being used.

Re:SETI

[Danathar]

You know of course you don't HAVE to use Fedora to run Condor. Condor will run under different OS's.

As for the benefit, let me give you an example of how Condor is being used. The Open Science Grid is a collection of Virtual Organizations that have contributed their clusters for use by Scientists who need them. So for example, when the massive data starts coming from the Hadron accelerator looking for the mythical Higgs particle, Fermilab will be able to use tens of thousands of computers all over the world that these organizations have contributed.

I have 15 Dell GX280's that were going to be scrapped and are now happily doing work. Sometimes it's biology experiments, other times physics, and when not running BOINC runs either einstein@home or Rosetta@Home.

Re:SETI

[Danathar]

Well, obviously it's your choice, but I stress it's not an either/or proposition.

The BOINC people have partnered with the Condor people because quite frankly many of these clusters DO sit idle a lot of the time. When they do, Condor can fire off BOINC to run in the background. My systems run 24x7 at 100%. If not running some job submitted by some scientist somewhere (and it's pretty flexible) then running rosetta@home or Einstein@home or whatever.

Note the systems I run are Dell GX260's, 270's and 280's which are on the order of 4 years old+ and they are used quite nicely.

An Athalon 2000+ is an older processor, but it's still valuable as a resource.

BOINC - a whole infrastructure

[DrYak]

I am all for open source, but there are some better places to donate some spare cpu cycles

And they are thing like BOINC [slashdot.org] which are complete opensource infrastructure for distributed computing, which are cross platform and feature lots of project you can pick from to contribute your spare time (among other, the original SETI, but also dozens of bio-medically related ones which will also have a similarly more close impact for humans as folding at home is).

Re:SETI

[somersault]

Nightlife is usually what you do before you boink.

It's not the idle capacity I'm worried about

[Nursie]

It's the not-so-idle electricty bill that'll turn up when I let people use my PC's spare cycles all the time.

That's why it's off, in stand by or auto throttling the processor. That's why letting people use your "idle" cycles is not as simple a charitable proposition as it sounds.

Re:It's not the idle capacity I'm worried about

[abolitiontheory]

What's the actual difference in energy costs, though? Not saying you're stupid or selfish for not donating, just interested in the real figures, if you've got any. I throw my system into hibernation most nights, and try to turn off the monitor at least when I go away for a couple hours during the day. What have you found your general savings to be?

Re:It's not the idle capacity I'm worried about

[Hairy Heron]

In my experience it's around 5 dollars a month more to run my computer all the time rather than shutting it down or putting it into hibernation at night.

Re:It's not the idle capacity I'm worried about

[somersault]

Presumably it was just idling when you left it on overnight though - what would the costs be if it was run flat out all the time? I leave my PS3 running folding@home all the time, I'm not too fussed about the extra cost - have done over 100 units now after having it for just over a month. If it were in my bedroom rather than the living room then I'd probably switch it off in the evenings because of the fan noise (it's not bad, but it's not silent either).

Re:It's not the idle capacity I'm worried about

[PitaBred]

So, $5 a month to help understand proteins better? It may be tax deductible if you really work at it, but it's still charity I can get behind

Re:It's not the idle capacity I'm worried about

[Nursie]

Real figures?

I don't know, but given that people have PSUs rated from 250W - 1KW these days, I would have thought fairly significant, assuming a pretty high utilisation of "spare" cycles.

I know we've managed to cut our electricity bill in half lately by moving to energy saving bulbs and making sure we actually switch stuff off at the socket when it's not in use.

Also, there's that whole "not using more than you need" thing to do with electricity having to come from somewhere, and that simewhere usually being a source of CO2 and other nasties.

Re:It's not the idle capacity I'm worried about

[abolitiontheory]

... making sure we actually switch stuff off at the socket ...

I keep trying to think of a funny and poignant way to point out that we Americans don't have the slightest notion of this concept, because its not built into to our electrical system. I'm sure you could get switches at the sockets if you intentionally looked for them, but I was 21 before I ever knew of this concept, from going over to England to visit family. It's one of those small details that sticks in your head, kind of like slang words or Cadbury chocolate. American chocolate is rubbish.

Re:It's not the idle capacity I'm worried about

[Nursie]

Come to think of it, you're right, I haven't noticed power switches on power sockets whenever I've been in the US (quite a lot over the years). Guess you don't notice the absence of little things like that.

Ghirardelli do an ok bit of chocolate. OTOH, Hersheys is like some sort of brown soap.

Actually unplugging everything would be more hassle, having the switch right there on the socket is a good thing, IMHO. I'm not obsessive about this and have a home server and a router that are UPS'd and on all the time. If anyone can get any use out of the spare cycles on my 266MHz ARM box then they're welcome to 'em...

Re:It's not the idle capacity I'm worried about

[hab136]

It's common on new construction in the US to have one light switch per room tied to an outlet. It's not common to have every outlet be switched.

Re:It's not the idle capacity I'm worried about

[xaxa]

In the UK (almost) every socket has a switch on the same plastic panel. http://en.wikipedia.org/wiki/Image:Uk_13a_double_socket.jpg [wikipedia.org]
It's possible to get sockets without the switch, but they're much less common (in a house you might find them in the kitchen behind fridges etc).

"Stand-alone" switches (by doors etc) only ever control wired-in lights. The first time I visited a house in the USA I was confused about how I turned off the TV when I walked into a room and hit the switch by the door. That's not possible in the UK. I can see the American system can be useful though.

Re:It's not the idle capacity I'm worried about

[analog_line]

A power strip with a switch does the job just as well as a switch on a socket. In fact, that's what I thought he meant until I read your post.

Re:It's not the idle capacity I'm worried about

[thePowerOfGrayskull]

I think you meant to say that "I personally didn't have the slightest notion of this concept before I went to England". In all the places I've lived, there were several receptacles that were controlled by wall switches. Considering that I've lived in six different states, I'd say that's a fairly reasonable sample size for one person.

Re:It's not the idle capacity I'm worried about

[drew]

It depends where you live. In most American houses built since at least the late 70's or so, you'll typically find at least one switched outlet in each room, because when building houses went from being a one at a time custom thing to a neighborhood at a time thing, developers decided that they didn't want to have to choose light fixtures for each house, and would rather save the time and money by having the light switches control an outlet that the eventual homeowner could plug whatever kind of light they wanted into.

Of course, they aren't on every outlet, and many Americans don't consider these switched outlets to be a feature, seeing as they are primarily associated with apartments and cookie cutter homes. I know in my house we've gradually been replacing them with hardwired fixtures as we do our various home improvement projects. But I'd be very surprised if the majority of Americans aren't at least vaguely familiar with the concept of switched outlets. At the very least, pretty much anyone that lives in a large apartment complex, condo, townhouse, or single family house in a (relatively) recently developed neighborhood certainly would be.

Oh, and I think you meant to say that American milk chocolate is rubbish. There are American companies that make perfectly good dark chocolate (assuming it's real dark chocolate that they are actually willing to put a %cocoa on the label. I hardly consider anything Hershey makes to be "dark".)

Re:It's not the idle capacity I'm worried about

[xaxa]

The grandparent should have included a picture:
http://en.wikipedia.org/wiki/Image:Uk_13a_double_socket.jpg [wikipedia.org]

Now you know what we're talking about :-). All UK sockets are like that (and some other countries).

Re:It's not the idle capacity I'm worried about

[drew]

I figured the implementation was different than what I was describing, but my point remains that, unlike the GGP, most Americans shouldn't be that utterly confused by the concept of turning off an outlet.

I hadn't the slightest idea what a UK power outlet looked like until just now, but I've been seeing / hearing people use the phrase "switching things off at the outlet" or similar for probably two years now, and I don't think I was ever once confused as to what they meant when they said it. So saying that Americans "don't have the slightest notion of this concept" seems a bit much to me.

Re:It's not the idle capacity I'm worried about

[Nursie]

I'm sure I've seen them in other places...

Maybe Singapore (OK, so they use british standard sockets).... Hmmm.

Re:It's not the idle capacity I'm worried about

[farmerj]

Yep the Irish power sockets are the same as the British ones. A lot of our fittings and standards are based on the British ones.

The main exception I can think of offhand is the phone sockets, the British ones use a BT socket [wikipedia.org], while the Irish ones use a RJ11 socket [wikipedia.org].

In the past this has caused a wee bit of consternation, as many retailers that service both the UK and Ireland sold phones with BT sockets in Ireland, leading to many WTFs as people tried to connect them :-)

Re:Switched Outlets For a Switch

[FurtiveGlancer]

Allow me to add Qatar, Kuwait, Iraq, Iran, Afghanistan, Kyrgyzstan, Italy and Germany to the list of places where switched outlets are used. I've seen the switches primarily on high current outlets for major appliances, such as washers and dryers.

I'd postulate that the switch at EVERY outlet in the UK may be related to WW II air raids and blackouts.

Re:It's not the idle capacity I'm worried about

[ozbird]

Actually, no-one other than the British do, AFAIK.

Australia has switched wall outlets; New Zealand shares the same plug/socket format, so they probably do too. It might have something to do with having 240V mains power (vs. 110V) and/or more stringent safety regulations e.g. we have earth pins on all outlets too.

Re:It's not the idle capacity I'm worried about

[Blkdeath]

Actually, no-one other than the British do, AFAIK. Haven't seen a switch in a socket in Portugal, Spain, France, Germany, Italy or the US. Maybe Ireland?

New Zealand does (Thanks, Discovery Channel!)

Re:It's not the idle capacity I'm worried about

[Angostura]

English and American chocolate is very different for historical reasons.

Historically, making milk chocolate was beset by one problem: getting fresh milk to the plant. Before refrigeration and railways it was nigh on impossible. The two countries had different approaches: in the UK sugar was added to the milk as a preservative, in the U.S, the milk was essentially allowed to sour.

By the time refrigeration arrived, the nations' tastes were set. The U.S continues to make much of its chocolate with soured milk, while the UK has very sweet milk chocolate.

So when as an Englander I first tasted a Hershey bar the overwhelming impression was of off-milk or vomit. This was, ummm... offputting.

But it is just what people are used to.

It's not the chocolate I'm worried about

[mi]

English and American chocolate is very different for historical reasons.

My point was not the differences in chocolate, it was the declaration of all American foo as "rubbish" and the subsequent high moderation...

So when as an Englander I first tasted a Hershey bar the overwhelming impression was of off-milk or vomit.

Come to think of it, I have a very vague recollection of how vomit tastes like... Always having it fresh in memory must be an English thing ;-) Sorry, could not resist.

Re:It's not the idle capacity I'm worried about

[UncleTogie]

As a US military brat, I've got to agree. US chocolate is the pits. Luckily, the local supermarket now carries the Milka brand I got hooked on when I was in Germany...

Real numbers

[BLKMGK]

I use the newer 80+ rated PSUs and I don't oversize them like so many others do. My desktop machine AND a server that also has an 80+ PSU in it (and 10HDDs) together use just about 300Watts as measured by my Kill-a-Watt device. That's not an insignificant amount but that was also with all of my drives spun up - normally drives not in use goto sleep (unRAID).

The PSU ratings of those two machines together are probably somewhere right around a kilowatt and yet I use a fraction of that at full chat. My desktop has a 45nm C2D (E8400) clocked to 3.8Ghz, an 8800GTS (die shrunk too), a single HDD, multiple cooling fans.

My point is that just because a PSU is rated for something doesn't mean it's going to be using that even when you have fairly thirsty components onboard - using the rating is a bit misleading as it's a maximum output. The fact that I use highly efficient supplies helps a great deal, they don't cost much more. My power bill isn't insignificant mind you but these aren't the only two computers I run either :-)

Re:It's not the idle capacity I'm worried about

[Idbar]

Well, as you can use your CPU cycles to help research, I guess you can also help by not wasting energy too. Both are common efforts and achieve better results as more people join. If blackle [blackle.com] has any numbers that only by changing the background (I have all my windows with dark background anyways, just because I don't get so tired of my eyes at the end of the day), you can help. I think, that as a common effort to achieve different goals, you are free to chose either one, independently on how much it cost to each individual.

Re:It's not the idle capacity I'm worried about

[Nursie]

Err, blackle is nonsense in the modern world, IMHO.

LCD/TFT screens don't work that way. There is a bright light that's always on, and the colours and darkness come about by blocking portions of said light, not by generating more of it.

Of course, once OLED comes in that'll change again.

Re:It's not the idle capacity I'm worried about

[Idbar]

Yes I was just setting an example. There are several places I know they still use CRTs, and I'd guess in that case it makes a little sense to use black. But if by turning your computer off, you are completely avoiding consumption, then makes much more sense, than any pseudo-screensaver.

Re:It's not the idle capacity I'm worried about

[milsoRgen]

hell my machine still has just CD-RW, but it has floppy drive)

I love that, "but it has floppy drive" part.

Re:It's not the idle capacity I'm worried about

[lubricated]

I've wondered about this. If the light is always on and you need electricity to not let part of it through then white should in theory waste less power than black. Also setting your backlight lower during the screensaver would help too.

Re:It's not the idle capacity I'm worried about

[nine-times]

Also setting your backlight lower during the screensaver would help too.

Or turning your screen off (or "power saving mode") instead of going to screensaver. Screensavers also don't make a ton of sense in the modern world. Burn-in is much less of a problem on modern monitors, and computers can power down the screen instead of going to screensaver.

In truth, I still use a screensaver as a sort of warning system. The screensaver comes on 3 minutes before the computer sleeps, so if I want to keep it from sleeping, I have a couple minutes to stop it. But if you're using some kind of cool 3D screensaver, I'd guess that it might use more power to display it than to just keep showing your desktop.

Re:LCD/TFT Screens Don't Work That Way

[FurtiveGlancer]

Fpanels turn off the backlight after a period of inactivity, typically 15 minutes. They must to earn an energy star symbol.

Re:It's not the idle capacity I'm worried about

[Matt Perry]

Err, blackle is nonsense in the modern world, IMHO.

Maybe so, but they did cite their source, which is a paper from the Lawrence Berkeley National Laboratory [lbl.gov] that supports their claim. See pages 15 through 19 of the paper.

Re:It's not the idle capacity I'm worried about

[Waffle Iron]

What's the actual difference in energy costs, though?

I just hooked a Killawatt to my Athlon 64 X2 4800+ system. Idle, it uses 67 watts at the wall outlet. Simultaneously transcoding two videos with mencoder reads 130 watts.

If this runs 24x7, the extra 63 watts would use 1.5 KwH per day, which would cost me $71 per year with my incremental electricity cost of about 13 cents per KwH. That costs almost as much as a subscription to Netflix.

Another consideration is that when idle, the system is almost silent. Under load, both the power supply fan and CPU fan crank up and get rather loud.

Re:It's not the idle capacity I'm worried about

[swillden]

You can go for a happy medium. Configure your system to keep the processors scaled back to minimum clock speed and then use the idle cycles. The power consumption will go up as compared to a completely idle system, but not very much, and you'll still have a lot of cycles to donate. Also, your machine should be able to sustain that load without speeding up the fans.

Re:It's not the idle capacity I'm worried about

[Waffle Iron]

I just tried your idea. Forcing the system down to 1GHz from 2.5Gz causes it to use 13 additional watts instead of 63 extra watts while under load. The frames-per-second performance scaled with the clock speed at 40%, while the power increment was only 20% as much as full speed. So on this test low speed was twice as "efficient" at producing results per additional watt. (Of course, that assumes the system was going to be left turned on anyway. If not, it looks like full speed would get through a given number crunching workload with the lowest overall energy usage.)

Re:It's not the idle capacity I'm worried about

[Waffle Iron]

It stayed quiet. The fans only seem to have a few discreet settings, and they didn't step up for that amount of power.

Re:It's not the idle capacity I'm worried about

[milsoRgen]

Yes but what is your video card? I have a ATI x1800 XT that runs hot even when idle, I've often wondered how much it has increased idle consumption.

Actual energy costs

[jgoemat]

What's the actual difference in energy costs, though? Not saying you're stupid or selfish for not donating, just interested in the real figures, if you've got any. I throw my system into hibernation most nights, and try to turn off the monitor at least when I go away for a couple hours during the day. What have you found your general savings to be?

A modern dual core processor can use about an extra 100 watts of energy when processing than when idle. This is from using a watt meter on a few computers of mine and checking it out. Shutting down or hibernating will save you 200-300 watts total I'm guessing. Personally I have a couple of services running on my computer all the time so I can't shut it down completely, plus I like being able to just turn on the monitor and start working/gaming/surfing. So if you are going to leave your computer running you can save 1 kilowatt-hour every ten hours just from the extra power pulled to do the processing versus having the CPU usage low. Prices [doe.gov] vary, but if your CPU usage totals an extra 20 hours per day then that's an extra 2 kilowatt-hours per day or about 20 cents, totalling up to $75 or more a year in electricity bills. That also increases carbon emissions if your electricity comes from fossil fuels.

rough estimate ...

[Lazy Jones]

Wikpedia [wikipedia.org]: Microwulf, a low cost desktop Beowulf cluster of 4 dual core Athlon 64 x2 3800+ computers, runs at 58 MFLOPS/Watt.

BOINC [berkeley.edu] does about 1200 TFLOPS (= 1,200,000,000 MFLOPS) atm.

=> BOINC probably burns around 20MW (assuming that the power used is directly proportional to the CPU time used even if it isn't 100%, which is wrong but an upper bound and probably not very far off).

1 KWh electricity = 0.43Kg CO2

=> BOINC generates 8.6 tons CO2 per hour or about 3100 tons/year (correct me if I'm wrong, I might be a few orders of magnitude off). That isn't very much compared to the 6b tons emitted by the USA anually, but still a waste...

Re:Energy costs

[Nursie]

And how about compared to turning it off when you're not using it?

"Under these assumptions, running BOINC costs about $3/month more than leaving your computer on but idle, and about $8.80/month more than leaving it off all the time.

There may also be an environmental cost. If your electricity is produced by burning fossil fuels, the extra electricity usage produces greenhouse gases that contribute to global warming. If this is the case, we recommend that you not leave your computer on just to run BOINC, or that you reduce your overall energy use to compensate. "

Re:It's not the idle capacity I'm worried about

[Yvanhoe]

Don't see it as free, see it as a really easy way to give some money that you know will go into CPU cycles quickly and efficiently.

And to donate your company's money as well ;-)

Re:It's not the idle capacity I'm worried about

[nitehawk214]

Don't see it as free, see it as a really easy way to give some money that you know will go into CPU cycles quickly and efficiently.

And to donate your company's money as well ;-)

In that case, you are better off switching your computer off and sending the organization a check for whatever it would cost to run your computer at max load 24/7. Not only will you give them a much more useful work for the dollar (if the million nodes they wanted instead just have them just one dollar, they could easily buy a supercomputer), it will also save the environment by not using electricity by a million inefficient home desktop machines.

Re:It's not the idle capacity I'm worried about

[Yvanhoe]

You mean, write the check, put it in a letter and put a stamp on it ? Of course one million people will not donate one dollar to this project, but they will more easily give 2$ in electricity money.

Re:It's not the idle capacity I'm worried about

[shawn(at)fsu]

not as simple a charitable proposition as it sounds.
Would the fact that it's costing you more money than you thought make it more charitable?

Re:It's not the idle capacity I'm worried about

[Nursie]

Weel, that would depend if it's just the money or the associated energy use and atmospheric pollution that's on your mind.

Re:It's not the idle capacity I'm worried about

[compro01]

to provide some numbers:

According to my UPS, my computer (3.0ghz core2 duo E6850, 2GB ram, 8800GTS 640MB, 500GB hard drive, including modem and switch) consumes about 180W at idle (monitor/speakers/etc. off, torrents running). Running FaH (same as before, including torrents, but with FaH running on both cores), it sits at about 220W. Powered off, it registers at 5W. Running flat-out, it registers about 350W.

Assuming it's running flat out 8 hours a day, that leaves 16 hours of off, idle, or FaH each day.

Price of electricity here is 6.5 cents per KWhr, so that gives us this for the basic monthly cost:

8*0.35=2.8KWhr = 18.2 cents per day or $5.46 monthly

Now for the other 16 hours a day :

off : 16*0.005=0.08KWHR = 0.5 cents per day or $0.15 monthly
idle : 16*0.18=2.88KWhr = 18.7 cents per day or $5.61 monthly
FaH : 16*0.22=3.52KWhr = 22.9 cents per day or $6.87 monthly

total costs :

off : $5.61/month
idle : $11.07/month
FaH : $12.33/month

Re:It's not the idle capacity I'm worried about

[CCFreak2K]

That's why I run it on my shell server out in some datacenter instead. We pay the same no matter how much/little we use of ANYTHING on the server, so we might as well make use of it.

Why Fedora?

[sysusr]

If they settled for Windows, the sheer volume of available machines would far outweight any (probably minor to begin with) advantages to using Linux.

Re:Why Fedora?

[sysusr]

Sorry, I'm an idiot. Please ignore me while I sit in the corner and watch...

It's a bit nebulous

[Silver Sloth]

Err... I've read TFA and all I can see is that some guy would like to use spare Fedora cpu cycles for some sort of project but he doesn't know what and he's not really sure how. My immediate response is come back when you've got something concrete.

Re:It's a bit nebulous

[morgan_greywolf]

Err... I've read TFA and all I can see is that some guy would like to use spare Fedora cpu cycles for some sort of project but he doesn't know what and he's not really sure how. My immediate response is come back when you've got something concrete

Hmph. Sounds just like a PHB when they propose a new development project. "Well, see, we want to use [ SAP | Lotus Notes | Teamcenter | other complex technology here ], but we're not really sure how we'd use it. For fsck's sake, if you don't already know HOW you would use something, you probably DON'T NEED IT!

Re:It's a bit nebulous

[Blkdeath]

Hmph. Sounds just like a PHB when they propose a new development project. "Well, see, we want to use [ SAP | Lotus Notes | Teamcenter | other complex technology here ], but we're not really sure how we'd use it.

For fsck's sake, if you don't already know HOW you would use something, you probably DON'T NEED IT!

As the old addage goes; when the only tool you have is a hammer suddenly every problem looks like a nail.

Re:It's a bit nebulous

[ronaldb64]

In other news, a new botnet was discovered. It consists of close to a million computers, all of them running Red Hat Fedora.
The new botnet for Linux is seen as proof that Linux is threatening the monopoly of Windows in more than one area.

I have a better name ...

[Lazy Jones]

how about calling it "red computing" to remind people of how much energy it'll cost them. On modern computers, you have roughly 20-100W difference between idle/working CPUs.

Re:I have a better name ...

[Ngarrang]

If we would just build more power plants, this wouldn't be a problem. I need more power for my cluster, anywho. And, if I ever complete my cluster powered by 1000 Mini-ITX boards, watch out world!

Bwuhahahahahahaha!

(This message has been brought to by Pave The Planet.)

Re:I have a better name ...

[Lazy Jones]

That's the right way really... build more power plants (preferably solar power plants), then if there is spare energy at peak production times, we may use it to trigger some grid computing by otherwise idle processors. Nothing wasted that way...

Re:I have a better name ...

[CCFreak2K]

According to my color wheel, the opposite of green (lime green, anyway) is magenta. Let's call it magenta computing.

Already done by others

[pwilli]

BOINC [berkeley.edu]

is a client that allows you to choose out of many projects like Folding@home or SETI. The client also runs on Windows, Linux and MacOS without problems.
There are many configuration options available to control the amount of CPU-power, cores, hard-disk space, RAM, the times it runs, how it should behave is someone else is using the system, etc. and the best is, anybody could set up a project that uses the client (although you'll probably have ahard time getting people to choose your project if it isn't something very interesting).

Check it out!

Isn't just this Boinc?

[nweaver]

The Seti-at-home crowd, long ago, realized that it was more than Seti@home, thus created BOINC [berkeley.edu]. So whats new here?

Has he not heard of Boinc?

[Bazman]

http://boinc.berkeley.edu/ [berkeley.edu]

  "Use the idle time on your computer (Windows, Mac, or Linux) to cure diseases, study global warming, discover pulsars, and do many other types of scientific research. It's safe, secure, and easy"

  And you can do it NOW. With almost ANY computer.

He's either not done his research or he's an idiot.

Re:Has he not heard of Boinc?

[Bob Loblaw]

From reading the linked articles, it seems the main difference is the way work is managed.

With BONIC, clients redundantly get sent out chunks of work that get send out again if they expire without some response. So this can lead to really large and unpredictable lag times between work scheduling and work completion. Which is great for some tasks but not so great for others.

With Condor, from the Condor website under the clearly indicated link "What is Condor?", "Should Condor detect that a machine is no longer available (such as a key press detected), in many circumstances Condor is able to transparently produce a checkpoint and migrate a job to a different machine which would otherwise be idle." So this means that work units are done with some consistent timing at the expense of increased complexity.

This represents a pretty large difference in how the two systems operate. So before calling someone else an idiot for not doing their research, you should at least do your own or fall under the same label by your own criteria.

Re:Has he not heard of Boinc?

[Bazman]

As easy as 'apt-get install boinc'? Or you can click things in synaptic. Ubuntu ftw again.

Re:Has he not heard of Boinc?

[Blkdeath]

And there is now a BOINC package in newer releases of Fedora.

Fantastic!

Now, I've just got a couple glitches to work out in my Fedora 9 installation;

1. Video drivers
2. Wireless LAN
3. Getting KDE to function and allow access to all the Fedora software utilities properly
4. Allowing me to use groupinstall/groupremove for KDE/Gnome without mucking up my entire graphical environment
5. Providing me a KDE utility to graphically change video settings would be spectacular
6. Fixing the GNOME graphic settings utility to allow for advanced changes like video timings, output selection, etc. would be a real boon as well
7. Convincing the package manager to look at the DVD instead of the Internet every time I ask it to install another package

On the plus side; since my Fedora installation is all but useless, I might as well donate some spare CPU cycles to something. I've got lots of them; I can't use the computer to save my life

Re:Has he not heard of Boinc?

[SanityInAnarchy]

Providing me a KDE utility to graphically change video settings would be spectacular

Fedora doesn't do that? Really?

On Ubuntu, I just go to System Settings, Monitor & Display, and drag the slider. Granted, System Settings isn't in KDE3 by default -- but it's just an OSX-like frontend to all the KDE control modules. There is an older frontend that you should have somewhere.

Mac users can do something like this already...

[Dekortage]

Since Mac OS 10.4 and later come with Xgrid [apple.com] already installed, it's very easy for your spare processor cycles to be donated to science [macresearch.org]. A few clicks in your System Preferences, and you're done.

World Community Grid

[Luyseyal]

Personally, I prefer World Community Grid [worldcommunitygrid.org]. I've been a member of the Slashdot team there since 2005 sometime.

-l

Re:World Community Grid

[H0p313ss]

Cound not find info about system requirements (platforms)

Boinc [berkeley.edu] is the current client, requirements are here [berkeley.edu].

Re:World Community Grid

[Luyseyal]

I run it on Linux, though there is Mac and Windows support as well. It's just a BOINC process. I use the Debian packages and added the WCG project and voila, I'm crunching numbers for science and humanity.

There are some folks who argue saving the energy is better (if you can't afford it, I agree!). However, I think it would have been worse for these projects to have never existed and/or to have built their own massive server farms. Spreading the energy burden around the world sounds to me like a better proposition than, say, building another coal plant to supply their growing farms.

I have wind power. Natch.
-l

Re:World Community Grid

[Secret Rabbit]

It is assumed that you know that you run it through BOINC. Though that topic is in the "Getting Started" section in Help:

http://www.worldcommunitygrid.org/help/viewTopic.do?shortName=start [worldcommunitygrid.org]

Also, all the projects available is right on there front-page.

Can't know anything without registering, eh?

Che Fedora!

[WheresMyDingo]

Greedy capitalists, share your idle cycles! Power to the people!

BOINC != grid computing....Condor is

[Danathar]

To all you people saying "why don't you just use BOINC"

Why indeed? Why not use BOTH. (As Condor can be configured to use BOINC when it's idle)

With BOINC data is PULLED from them to you when YOU request it. In grid computing with Condor data is PUSHED to you.

Big difference.

my spare cycles?

[drew]

I prefer Electric Sheep [electricsheep.org].

Re:mount fsck fsck ...

[simong]

There isn't the computing power in the world for that...

Re:mount fsck fsck ...

[UncleTogie]

There isn't the computing power in the world for that...

Sure there is... and it's at Craigslist, [craigslist.org] of all places...

Re:In 3..2..1

[H0p313ss]

I love Nightlife...I love to boogie.

boogie woogie?

Re:In 3..2..1

[jav1231]

Offtopic!? :( Tough crowd.

Alicia Bridges? [wikipedia.org]

"So I Married An Axe Murderer?

Re:Not responsible computing

[Ramsés Morales]

Did you send this message from a cave, using no electricity?