Slashdot Mirror

← Back to Stories (view on slashdot.org)

Harvard, IBM Crunch Data For More Efficient Solar Cells

Posted by samzenpus on from the better-than-the-sun dept.

Nerval's Lobster writes "Harvard's Clean Energy Project (CEP) is using IBM's World Community Grid, a 'virtual supercomputer' that leverages volunteers' surplus computing power, to determine which organic carbon compounds are best suited for converting sunlight into electricity. IBM claims that the resulting database of compounds is the 'most extensive investigation of quantum chemicals ever performed.' In theory, all that information can be utilized to develop organic semiconductors and solar cells. Roughly a thousand of the molecular structures explored by the project are capable of converting 11 percent (or more) of captured sunlight into electricity—a significant boost from many organic cells currently in use, which convert between 4 and 5 percent of sunlight. That's significantly less than solar cells crafted from silicon, which can produce efficiencies of up to nearly 20 percent (at least in the case of black silicon solar cells). But silicon solar cells can be costly to produce, experiments with low-grade materials notwithstanding; organic cells could be a cheap and recyclable alternative, provided researchers can make them more efficient. The World Community Grid asks volunteers to download a small program (called an 'agent') onto their PC. Whenever the machine is idle, it requests data from whatever project is on the World Community Grid's server, which it crunches before sending back (and requesting another data packet). Several notable projects have embraced grid computing as a way to analyze massive datasets, including SETI@Home."

10 of 65 comments (clear)

  1. Also, SETI... by Anonymous Coward · · Score: 2, Insightful

    Nice of them to mention that SETI has also "embraced" this. It's only the largest and one of the oldest public projects to utilize distributed computing, having lead the way in the development and popularization of the technology.

    1. Re:Also, SETI... by jellomizer · · Score: 2

      The interesting thing about SETI it is one of those projects where the side effect technology is more important then their main research goal.
      Scanning Space Radio signals trying to find Extraterrestrial life is in itself rather futile goal.

      1. Radio Waves bounce off atmosphere. Sure it does get threw however most of our random singles from TV really get so bombarded by static that the waves data is nearly traceable due to the static from space. Even before we leave the solar system.

      2. Radio is only used for a short period of time. We have been using radio for about 150 years, and for long range broadcasting it is getting much slimmer. As we have built an infrastructure to send most data without radio, we use radio for short range broadcasts. So say we have about 500 years of useful radio useage. In terms of tracing a society that is like a blip. So you need to find the star, with a planet, with life, who is intelligent, has an instinct to build stuff, figured they needed radio, happen to run across similar experiments to discover it, had the engineering mindset to make something out of it, used it over the alternatives, and hasn't outgrown it. This is slim pickings. If you think about it, it was said the burning of the Library of Alexandria set mankind back 500 years. Who knows how things would be different.

      --
      If something is so important that you feel the need to post it on the internet... It probably isn't that important.
  2. Why Efficiency? by lobiusmoop · · Score: 3, Interesting

    I don't understand why efficiency is so important - $/W seems a much more important measure, given that arid land area is cheap and sunlight is free.

    --
    "I bless every day that I continue to live, for every day is pure profit."
    1. Re:Why Efficiency? by kcbnac · · Score: 4, Insightful

      Because if you can build one at 3% and one at 9% but all other costs being the same, you build the 9% one. They're figuring out which is the most efficient so they know what order to look at capabilities/options on. Start with the most efficient, and work your way down the list until you find one that meets the other criteria.

    2. Re:Why Efficiency? by Waffle+Iron · · Score: 3, Interesting

      LOL. Try actually computing how much land area is required to cover 1% of US electrical baseload given 20% solar panel efficiency and 600W - 1000W incident solar radiation per square meter on a perfectly sunny day in May (e.g. most optimal time of the year.)

      You should quickly come to the realization that solar is a complete boondoggle, even at 100% panel efficiency.

      Annualized average US power = 440GW. Your 1% of US power is 4.4GW. At 20% of 800W/m^2, that's 6800 acres, or about 10 square miles.

      You seem to have a problem with that?

      IIRC, about 1/6 of the 90,000,000 acres of corn grown in the US goes to silly schemes to make ethanol, which is 2000 times as much land as your example. That seems like a prime candidate to replace with far more efficient solar panels, especially if the areas with the most dire depletion of aquifers are reallocated first.

  3. Am I being too cynical? by tippe · · Score: 2

    a 'virtual supercomputer' that leverages volunteers' surplus computing power

    The first thing I thought when I read that was: "Fools. You're wasting your own energy to fund somebody else's patent portfolio (and wallet)." The idea that this might be a good idea or could forward some facet of science or could make the world a better place didn't even occur to me. I'm getting to be too cynical I think...

    1. Re:Am I being too cynical? by RicktheBrick · · Score: 2

      There are several problem with relying on volunteers. First one does not know for certain if the volunteer is going to return the finished work unit. So one must put a deadline on returning it. If one sends a work unit to one volunteer and they fail to return it before the deadline than one will have to send it to another and give another deadline. Since this pattern could repeat itself it is better to send the same work unit to more than one volunteer to ensure at least one is returned. In fact it is better when more than one volunteer returns the same work unit since than one can compare the results to ensure that they are the same. There are reasons why a incorrect work unit will be returned so one must be sure that they are correct. This clean energy project is very computer intensive. There are work units that require 40 hours or more. These units should not be given to someone who is just computing during idle times. Since if one has only one or two hours of idle time a day than the work unit will require between 20 to 40 days to complete. That is a long time to wait especially when one does not know for sure if it will ever be returned. When more than one have to do the same work unit than that means efficiency is at least cut in half. Now supercomputers are getting both cheaper and more efficient. They now are using a lot less power per calculations than any home computer. I have read that the best now can do a Giga flop per watt. The big question is "How much more energy will be used by sending the work to inefficient and multiple home computers than would be used by an efficient supercomputer?" One must also pay for the bandwidth to both send and receive the work units and the servers and the power to run the servers. I believe that the cost of purchasing and running a supercomputer is probably less than using volunteers. Except by using volunteers one using their electricity and requiring them to pay for it. Now if one could get the volunteers to donate money that they would pay for electricity and wear and tear than the use of a supercomputer would make a lot more sense. Now I say this and look at IBM's record since they do have computers that are doing the work units. In fact they lead all the volunteers by a large margin(over 45 million results to around 24 million for second place). I would think that IBM would use its money in the most efficient manner so maybe my concerns are not well founded. The federal government does give away time on supercomputers to scientist. Whoever is in charge must think this project is not worth the time. I would think that IBM could give enough prove of the value of this project that either the federal government would give a grant or that enough money would be provided by investors. Is the chance of success so low that IBM must require the donation of computing time to get this started? Will there be a return of any money to the volunteers if this is a success? Will a 1% take over the project once it succeeds and reap the rewards of all the volunteers?

  4. Summary by lazarus · · Score: 2

    I've been on Slashdot for (what seems like) forever, and this article summary is probably the best I have ever seen. Well done Nerval's Lobster!

    Sadly I have nothing intelligent to say about the content.

    --
    I am not interested in articles about life extension advancements.
  5. Public Good, Or Profit? by Jane+Q.+Public · · Score: 2

    "Nice of them to mention that SETI has also "embraced" this. It's only the largest and one of the oldest public projects to utilize distributed computing, having lead the way in the development and popularization of the technology."

    Also, SETI@home and Folding@home, etc., use the BOINC infrastructure, not IBM's. You can be fairly certain that BOINC projects will not be used for corporate profit unless it's a corporation that is sponsoring the project.

    Not necessarily so, using IBM's infrastructure. When have they ever done anything that wasn't for corporate profit. Hell, they even shipped Hollerith-type machines to the Nazis during WWII to help keep track of the prisoners in the concentration camps.

    (And before you argue: YES, they did. It is solidly documented and there are records indicating that Thomas J. Watson personally knew about it.)

  6. Open Access by godel_56 · · Score: 3, Informative

    It's a condition of entry that all the results derived from grid computing work on World Community Grid, of which CEP is a sub project, must be made freely available to all researchers. That said, someone will have to go on and commercialize the work and so make a profit somewhere, but at least everyone gets an open go at it.

    IBM do not own the results of this research, they're just sponsors of the central hardware and storage, and help with initial programming and set-up.

    CEP is the only one of the World Community Grid projects that I don't crunch for as it has fairly onerous data transfer and computing requirements. It's a bit of a PITA.