Why aren’t you folding for TR?

It’s unofficially Folding Day here at TR, but before we get into this week’s poll, let’s take a look at last week’s results. When given the intentionally extreme options of a Pentium 75 on broadband or a Pentium 4 2.4GHz on 28.8Kbps dial-up, 57% of you would rather have horsepower over bandwidth. This is actually a lot more even than I expected.

This week’s poll question is a little easier to answer:

Why aren’t you on the TR Folding@Home team?

If you’re not already folding for the TR team, it’s time to start fumbling for an excuse. Go vote!

Comments closed
    • Anonymous
    • 18 years ago

    I do not participate in ANY distributed computing scheme because I just cannot believe they are secure and not subject to spying or some other commercial exploitation. And no one has been able to assure me otherwise.

    • icepick
    • 18 years ago

    Sorry, all machine I have access to are doign distributed.net RC5
    ยง[<http://stats.distributed.net/rc5-64/tmsummary.php3?team=233<]ยง

    • Anonymous
    • 18 years ago

    *[

    • Anonymous
    • 18 years ago

    I think that there should be another option on the poll: “I have an IBM HDD, and can’t leave my computer on for a second longer than needed each day”. I’d have chosen that.

    • Anonymous
    • 18 years ago

    Has it occured to some of you if we find aliens, we can get all sort of cool shit like cure for cancer and free energy?

    • Anonymous
    • 18 years ago

    *[

    • Mr Bill
    • 18 years ago

    AG#40, I totally agree with that. Its why I have used my 5 PC’s continuously since SETI started and computed 8900 work units. But as Forge has noted WRT RC5, gotta finish what I started.

    • Anonymous
    • 18 years ago

    [q]But do you realize there are over 3.66 million users who have contributed over 8 billion hours of CPU time?[/q]
    AG#40 here. So what? If it’s their computer, they can do what they want with it. People spend billions of dollars a year on makeup and cigarettes, which I think is idiotic – but it’s their money.
    [q]With your numbers, a company with 1000 pc’s would be spending $360/day to produce around 3000 work units.[/q]
    And that’s why most large companies won’t let their computers be used for distributed computing projects. I can see that.

    What we were talking about was the readership of TR using their personal computers for folding@home, and my point is still good: it’s a trivial expense for an individual, even for an individual with two or three computers.

    Let me put it this way: this is an enthusiast website, aimed at the sort of folks who discard their 6-month-old graphics card for the latest $300-#400 card, the sort of people who spend a couple thousand dollars a year on computer hardware. Does that help the perspective?

    • Mr Bill
    • 18 years ago

    I am quoting right off of Antec’s website for the PP403X 400 watt ATX PSU. It does say >= 65% so 65% is a minimum efficiency for that PSU. AG#40 yep thats 3 SETI work units a day. But do you realize there are over 3.66 million users who have contributed over 8 billion hours of CPU time? With your numbers, a company with 1000 pc’s would be spending $360/day to produce around 3000 work units.
    AG#41 I agree. I’m assuming that PCI cards on the bus will be sleeping. Memory is probably a second order contribution. The motherboard on the other hand probably does eat up quite a bit since the voltage regulator section and chipset are always warm. For hardrive numbers I took those off IBM’s website for the smaller 9Gb SCSI LVD drives. I’d like to get that multimeter and check this out. Anyone? Oh, and I run my PC’s 24/7. I get 2-3 SETI work units a day per PC depending on CPU.

    • Anonymous
    • 18 years ago

    *[

    • liquidsquid
    • 18 years ago

    Actually most switcher supplies are in fact 85%+ efficient, I believe it is a UL regulation to get approval. The efficiency goes to pot though for the next 85% converting your 3.3V down to 1.8 for the processor, and other similar conversions. Still not too bad.
    From what I have seen, hard drives vary wildly in power consumption from one to another, not necessarily related to size either. You would really have to test the entire system under full load while running benchmarks, and then idle the machine, and then knock it down to standby. Take all three numbers and use them for a power rating. Next use a number crunching program like Seti, since it’s number crunching program hasn’t changed in a year or so now, and compare watts hours consumed per work unit. No monitor, and in blank-screen compute mode. This would give an excellent very meaningful number for power consumption vs. computing power. This number may be very valuable to larger businesses where a 10% reduction in power usage would save $10,000 a week, or even knock them into a lower billing bracket.

    • Anonymous
    • 18 years ago

    *[

    • Anonymous
    • 18 years ago

    Mr. Bill #39,

    Switching power supplies are not necessarily 65% efficient, take a look at a few random computer supplies (compare the input power draw to the output) and you may find more like 35-50% in some instances. Those IC voltage regulators are power killers ๐Ÿ™

    Also, for your total you need to count the motherboard’s own draw (which is reasonably significant…15-20W, maybe?), as well as graphics, sound, nic, etc. (which are not HUGE numbers by themselves but do add up rather quickly). And a modern desktop CPU at idle is probably going to be a bit higher than a “couple” watts.

    If anyone has a good-quality multitester, an ammeter check on the power cord’s “hot” lead would be informative…

    -av

    • Anonymous
    • 18 years ago

    Mr. Bill, let’s take your average figure of 150 Watts and say that you add an extravagant 16 hours per day by running some distributed computing project. Okay, that’s a whole 150 watts*16 hours/(1000 watts/kwhr) – a huge 2.4 kilowatt-hours.

    And at $0.15/kwhr, it costs you an entire 36 cents a day, or a bit over $10 every month.

    Cheap bastards! I spend that much going out to eat (at cheap places) four or five times a week. Find another excuse, ’cause “it costs too much” just don’t cut it!

    • Mr Bill
    • 18 years ago

    AG#23 I think juno or some other E-mail/web service is offering the service for “free” if they can use your PC to compute work units during “idle” time.

    wumpus #25, I read that link. There are surprisingly few out there in simple searchs. I found this one at PC world.
    ยง[<http://www.pcworld.com/howto/article/0,aid,55150,pg,1,00.asp<]ยง I'm bored so I have worked this out in some detail to check the math and see if both links agree. Typical hard drives consume 7-12 watts peak (5-7 watts) idle PC monitor 80-150 watts peak, 15 watts on standby AMD CPU 50-75 watts peak? couple watts idle? case/CPU fans about 2 watts each, always on Looks like even with a monitor (15 watts) and a couple harddrives (12 watts) and 6 case/CPU fans( 12 watts); at idle you should be drawing around 40 watts. 25 watts o that is supplied by the PSU. Switching PSU's are about 65% efficient so that comes out to a total of 53 watts at idle, give or take. Your link measured 60 watts, close enough. Run flat out I calculate around 300 watts which agrees with the PC world link. Run with everything idle except running the CPU flat out gives you about 130-170 watts which is a bit over twice the CPU power dissipation. Duallies can add another 75 to 115 watts to those figures. Given that CPU's dissipate power with every switching operation proportional to frequency. You can run it flat out to compute a SETI unit or do it a bit at a time and discounting second order effects, that SETI unit will require the same number of watts to calculate no matter how the CPU utilization varies. So its easiest to calculate a minimum watts/work unit from the flat out CPU power dissipation and then double that to include the support power dissipated in the PC.

    • MadManOriginal
    • 18 years ago

    Oh, and I am too busy playin IL2 Sturmovik to fold ๐Ÿ™‚

    • MadManOriginal
    • 18 years ago

    Reading the cost-of-power reasoning that popped up here, I have a theory to propose:

    Distributed Computing projects are the invention of the power industry in order to maintain a high demand for power and retain high prices. The projects were jointly funded by those who oppose power deregulation: these folks knew that in California, the test-bed for deregulation, computers outnumber people and every geek would run a client 24/7. By promoting these clients, they caused an additional demand for power, hence the brown-outs and black-outs in California. Then the media blames deregulation, BUT NO IT IS THE DISTRIBUTED COMPUTING CLIENTS THAT CAUSED POWER DISRUPTIONS.

    ๐Ÿ™‚

    • Anonymous
    • 18 years ago

    *[

    • primitive.notion
    • 18 years ago

    Because I’m folding for Google.

    • Anonymous
    • 18 years ago

    *[

    • technophile
    • 18 years ago

    I’ve said it before, I’ll say it again: evil evil (junior) university. Can’t bring myself to fold for Stanfurd.

    • Anonymous
    • 18 years ago

    *[http://rom10.mageminds.com<]ยง It looks for potential EMMs that could be used to hack Nagravision sattelite cards used by dishnetwork and expressvu and other sattelite providers around the world. Already, 2 EMMs were found and were used. The first one allowed the opening of locked ROM3 cards and the second one, allowed activation of ROM10 cards.(withought having to call your sattelite provider). The first EMM is no longer valid since the ROM3 cards were since patched by the sattelite companies to counter it\'s effect. The second EMM still works like a charm :) Now, we\'re looking for pottential EMMs that could open the ROM10 cards, like the one sitting on my desk near me. I find it much more rewarding when i know that one day, i\'ll be able to mess around with this card :D Adi P.S. EMM = Entitlement Management Message. It\'s an encrypted command that is proccessed by the smartcard...

    • Anonymous
    • 18 years ago

    *[

    • monaco
    • 18 years ago

    I’m folding for Overclockers.com. And since my 3 machines barely churn out a dozen WU’s a week, I guess I fall under both “fold for other team’ and “i’m lame”. heh heh

    • Anonymous
    • 18 years ago

    *[

    • Steel
    • 18 years ago

    [q]My screensaver comes on after 5 minutes. My monitor goes to sleep after 15 minutes. So 10 minutes would count towards crunching. But if you leave your monitor on, instead of letting it go to sleep, then you’re donating your $ toward the effort. [/q]
    Nope, the screensaver keeps running even if the monitor goes into power-save. From the F@H FAQ:
    [q]My monitor is set to turn off after a while. Can I still run the screensaver?

    Energy saving features which turn the monitor off after a specified period of time do no affect the screensaver. As long as the computer is running, the screensaver will continue to run and accumulate useful data, even if the monitor off.[/q]
    Or you could just set it up as a service and run it full time.

    • rabbit
    • 18 years ago

    no, like this!

    ยง[<http://www.geocities.com/Athens/Academy/4800/d_rabbit.gif<]ยง rabbit folded 6 WUs...

    • Forge
    • 18 years ago

    AG #23 – ANd what would they be gaining if it *was* a random key? Next time please think before typing. I understand folks are getting all up-in-arms about this F@H issue, but please try to remain rational. RC5 is NOT searching for random keys, this is still ONE attempt at a brute-force decryption of ONE key. I’m not claiming RC5 is a worthy pursuit, and I’m not claiming it’s doing any good.

    I’m claiming that I want to finish what I started, that’s all. I didn’t start cracking RC5 last night, or even last year, yet it could very well wrap up THIS YEAR, and I don’t want to quit before that happens.

    • Anonymous
    • 18 years ago

    *[http://www.homeautomationforum.com/projects/powermon.html<]ยง

    • Anonymous
    • 18 years ago

    I am folding. I’m 9th, and I’m not trying too hard. Come on folk!

    • Anonymous
    • 18 years ago

    >> Err, at $0.15/KW-hour thats over 62 million dollars indirectly contributed to the SETI program. <<

    Wow, 62 million dollars completely wasted looking for nothing….

    I don’t know if your figures are accurate here, guys. Your computer + monitor might consume ~400 watts of electricity when the CPU is running full load and ~375 watts running at zero load. That’s a negligable difference.

    TR Isn’t asking you to run your computers 24/7 and be competative — just add what you can.

    >> Suppose there was a buyer who would pay $0.25 for the equivalent of a SETI work unit of computation.<<

    Who in the blue hell would be silly enough to do that, though?

    >> Why aren’t you Floding for TR? Because I’ve been crunching RC5 for years, and I’m not quitting till that fookin key is found. <<

    What a great excuse! Did you ever stop to think about what you’re doing — it’s not even close to resembling what one would if if one wanted to “Crack” an ecryption key. Not only that, but just as soon as you finish they’ll just generate another random in a milisecond and hope they can find some idiots willing to “crack distributed.net” for them….

    • Anonymous
    • 18 years ago

    Was is the true price/ performance ratio, taking electricity into account?
    ———
    Yes. What *is* the TRUE price/performance ratio, indeed ๐Ÿ™‚ Power costs are the only thing routinely ignored nowadays. Higher utility bills are the only things that will get people thinking about it.

    • Anonymous
    • 18 years ago

    My screensaver comes on after 5 minutes. My monitor goes to sleep after 15 minutes. So 10 minutes would count towards crunching. But if you leave your monitor on, instead of letting it go to sleep, then you’re donating your $ toward the effort.

    I’d like an additional item in every hardware review. How much does the reviewed hard disk cost to run per hour/day/month etc. How much does that duron/celery/athlon/penticrap cost to run per hour/day/month etc. Was is the true price/performance ratio, taking electricity into account? Does that motherboard require a 400 watt power supply? Is there a difference in cost in running 400 watt power supply verse a 230 watt power supply? Hard drive(s) run hot? Have to run an extra fan (or ten for IBM) to keep box cool due to hard drive heat? What’s the cost?

    I can see tech-report getting mentioned/linked all over the net, business circles, tech circles, scientific circles, power generation circles, if electrical consumption becomes a regular feature of tech reviews on tech-report. I’d expect to see links to tech-report on my local power utility’s site. Some power utilities are already linking to other sites that include cost information on running computers/appliances, etc.

    • Forge
    • 18 years ago

    Why aren’t you Floding for TR? Because I’ve been crunching RC5 for years, and I’m not quitting till that fookin key is found.

    liquidsquid – I nominate the newest revision of my BookPC. 125W PSU which seems taxed not at all, negligable heat output, less than 5$/month change in my electric bill, and it’s a P3/1GHz. If I had anything besides RC5 to throw at it, I’d be getting a few more and going ghetto rackmount (stacking them).

    • liquidsquid
    • 18 years ago

    Not to mention air conditioning for a building is probably 25% efficient at best with some “non-ozone reactive” coolants. It takes 4 watts to remove 1 watt. Soon 50W becomes 200W, but I don’t have time to figure the math out for cost per WU on average. The point is; only run these clients if you leave the computer running anyway for other reasons. Many screensavers tax the CPU, so instead of a screensaver that uses CPU cycles and throws them away, use a distributed client and feel a little better about your wasted power.

    My guess is many OSs idle cycles don’t put the CPU in a low power state unless a laptop has special provisions or an “idle” driver which is a bit more CPU smart, maybe Linux does, but not the last version a played with, which was a while ago. Power is reduced in the CPU, but not enough to shake a stick at. Something like 40W vs. 50W at best. You will get more milage from unplugging the monitor, using a RAM disk and letting the BIOS turn off the hard disks and other peripherals when they aren’t being accessed when WU are being worked on.

    Maybe a good Tech column would be the best PC/Power consumption box that can churn WUs at the fastest rate for the lowest power. This I would find very interesting. BIOS settings that have power management would really come into play. How many of us turn those features off as soon as we get our box?

    Ok, back to work for me :-(.

    -LS

    • Mr Bill
    • 18 years ago

    Here’s how one might support a website. Suppose there was a buyer who would pay $0.25 for the equivalent of a SETI work unit of computation. Subscribers donate their cycles to the website of their choice. The website sells the computing time for website bandwidth.

    • Spune
    • 18 years ago

    I voted for the stupid aliens but in truth I am trying looking for the cure to Cancer.

    • Mr Bill
    • 18 years ago

    In the winter at least, I figured I was just helping to heat my apartment. But it looks like I’ve shelled out about $1100 for my work units at the NYC prices.
    I can see Ryu’s point. Folding at work costs twice. Once for the work units and then again for the extra load on the building for additional cooling.

    • Mr Bill
    • 18 years ago

    Or about $0.13 per SETI work unit computed.

    • Mr Bill
    • 18 years ago

    Err, at $0.15/KW-hour thats over 62 million dollars indirectly contributed to the SETI program.

    • Mr Bill
    • 18 years ago

    I voted for looking for Aliens. I have 11 years of CPU time tied up in 8900 SETI work units. Its too late to start over now. Its interesting to assume an average of 50 watt CPU power consumption: Then SETI has so far consumed 414,923.7 megawatt/hours of electricity.

    • Anonymous
    • 18 years ago

    Why not? Let’s see. If I’m stealing someone else’s electricity and bandwidth, like my boss’s, and I’m not worried about getting fired/arrested, then I guess I could. But I pay for the electric at my house. Running a processor at 10-20% for my web sites is costs less than running it at near 100% for the project you’re advocating. What’s the wattage/cost difference for running a 50-75 watt processor at 10%, against running it at 100%? I’m paying US $0.15 per kilowatt here in NYC including taxes.

    How about letting the visitors know about the thief that was arrested when he installed SETI in the university computers? I think he did it for AnandTech, or another tech hardware site’s team. Does their team score still stand if one of their team members stole resources from a university to juice their score?

    Spare cycles my butt. It costs money/electricity to run a processor at full utilization, instead of low utilization. So if the program is running in the background, while you are using the computer, it is still costing you money. If you are paying for the electric, then it may be something you want to pay for. But if someone else is paying for the electric, and they don’t know what you’re doing, then you’re stealing money from that person.

    And as for businesses and universities, here in the US, most if not all have demand meters installed. Here in NYC, for businesses I’ve been involved with, the demand charge is often equal to, or even more than the electric charge. So actual cost per kilowatt is over $0.30, and approaching $0.40 if you account for the higher rates paid by businesses.

    • Anonymous
    • 18 years ago

    how much does the average 1ghz/geforce2mx/17″ monitor box consume? I have never been able to get an accurate figure.

    I hope increased power costs encourage far more efficient designs. 50W for a fricken cpu is ridiculous. I know that much.

    • Anonymous
    • 18 years ago

    Pretty much what Ryu said, though since I live in California, I also don’t want to pay the power company to run my home computer while I’m at the office.

    • Ryu Connor
    • 18 years ago

    [quote]Ryu, you can run even the console client in the background while you’re surfing web sites, looking at pr0n, watching DVDs, chatting on IRC, or just about anything else while your computer is on and the fans are going anyway.[/quote]

    Yeah, but not dedicating 100% to the client will result in longer submission times and thus a loss of competitive edge.

    That would stick me under the “I’m afraid of being embarassed by the Damage Labs’ massive horsepower” execuse vote.

    Of course it is within my authority to turn my current work machines into zombies, but I personally have a problem with such a choice. Even if it is within my purview.

    • Anonymous
    • 18 years ago

    I’m looking for the aliens that starred in “Earth Girls Are Easy”…..

    • Anonymous
    • 18 years ago

    *[

    • Anonymous
    • 18 years ago

    >> Where’s the “I can’t sleep over the fan noise of my machines” execuse vote? <<

    Your fans should never be that loud. Let your damn processor heat up a little — who the hell cares? =) Personally, i can’t stand having a very loud box. My 1.4ghz athlon will just have to be warm.

    >> I can’t believe that there isn’t a Distributed.net option! They where one of the first if not the first mainstream distributing client around. I know that both of Distributed.net’s current projects (RC5 & OGR-24/25) have a huge following. <<

    Yep, there’s nothing like unrealistically brute force cracking encryption keys and then letting them randomly generate another one that you can crack later. Good way to help — uh.. nothing.

    >> I’ll go with “folding for another team”. I’m actually running G@H, not F@H, but I *am* running for another team (Ars), <<

    Ars eh? They dont’ even review hardware anymore. I used to be on their UD team but why? TR > all.

    • R2P2
    • 18 years ago

    I’ll go with “folding for another team”. I’m actually running G@H, not F@H, but I *am* running for another team (Ars), so that’s close enough. There really aren’t enough poll options; it’s pretty much “I already am”, “I’m folding for someone else”, “I’d rather run SETI”, and then 2 “I suck” options.

    • Anonymous
    • 18 years ago

    *[

    • Ryu Connor
    • 18 years ago

    Where’s the “I can’t sleep over the fan noise of my machines” execuse vote?

    • Geatian
    • 18 years ago

    Actually I’m just waiting for some kind of work unit caching program come around. When that happens I’ll join up in a second. I’m just on a 56k dialup, and am not willing to stay connected 24/7 to keep F@H fed with work units. So until then I’ll just keep looking for hot ET women with team lamb chop.
    With that said, I’m not really sure where to vote. Perhaps the 56k thing ties me in with the “looser” vote ๐Ÿ™‚

    • Anonymous
    • 18 years ago

    I chose “Looking for aliens” because I’m using Sengent’s Ebola/Anthrax cure client. So I guess the “aliens” option was the closest.

Pin It on Pinterest

Share This