Slashdot Mirror


Sun Grid Compute Utility

jbltgz writes "The Register is reporting that the long awaited Sun Grid Compute Utility has been opened to the public. Now you can run your CPU intensive jobs on a grid of AMD Opteron-based Sun Hardware for $1 per CPU per hour for a fraction of cost, in a fraction of the time."

185 comments

  1. Selling off CPU time... by ZzzzSleep · · Score: 5, Interesting

    How long will it be until botnet operators start up a similar service? Or am I out of date and they have already done this? Anyway kudos to Sun for offering this service.

    ZzzzSleep

    1. Re:Selling off CPU time... by cant_get_a_good_nick · · Score: 1

      They already do, botnets have been rented out for both Spam and DDoS attacks for some time. In fact, the price kept going down, since it was so easy to set up a botnet.

    2. Re:Selling off CPU time... by Kadin2048 · · Score: 1

      Rent a botnet? Why bother -- submitting stories to Slashdot has been free for years!

      --
      "Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
    3. Re:Selling off CPU time... by jb.hl.com · · Score: 1

      How long will it be until botnet operators start up a similar service? Or am I out of date and they have already done this?

      It's called Bonzi Buddy...like Seti@Home for spam :P

      --
      By summer it was all gone...now shesmovedon. --
    4. Re:Selling off CPU time... by joecr · · Score: 1

      I guess you haven't heard about b3d Projector. They would install it with Kazaa & then Brilliant Digital would sell you processor time. (This was back in 2001-2003 AFAIK. They may have cleaned up their act since then, but I doubt it.)

      I quick look at their site & they make no mention of doing this in the past. I don't know if they still do that. I had to use a coral cache to even pull it up because I have blocked the brilliantdigital.com domain in my router. I made sure I didn't download any of the software they have as I still don't trust them.

      Another option although I don't know if they would hijack your processor time was new.net. I don't trust them either. I had to use a coral cache to even pull it up because I have blocked the new.net domain in my router. I made sure I didn't download any of the software they have as I still don't trust them.

  2. POVRay by AKAImBatman · · Score: 5, Interesting

    I wonder how long it would take for someone to port the POVRay engine to Sun's grid? At $1 per CPU/hour, this could be a boon for amatuer 3D graphics designers and the Internet Ray Tracing competitors. Use low res renders during testing, then pay Sun $25 to get your high quality result back in 20 minutes rather than the next day. Could be a lot of fun. :-)

    Can anyone think of other good uses for the average (or not so average) home user? Perhaps new image compression formats that rely on Sun's Grid to get the best compression/quality tradeoffs through brute-force power?

    1. Re:POVRay by Umbral+Blot · · Score: 1

      Best use of Sun's grid: playing Duke Nukem Forever, thats why it has been delayed so long, honest

    2. Re:POVRay by Anonymous Coward · · Score: 2, Funny

      Well, now that you asked, you can use all that computing power to...

      Compile gentoo with KDE in only 20hrs

      Browse 10 pages in Firefox

      Run Windows XP Pro AND Notepad at the same time

      Get 20FPS in BF2

      Run a "Hello World" java applet

    3. Re:POVRay by Anonymous Coward · · Score: 2, Informative

      Actually someone from SGI ported POV Ray to MPI about 4 years ago:

      http://www.verrall.demon.co.uk/mpipov/

      There is also a PVM version as well.

      http://www-mddsp.enel.ucalgary.ca/People/adilger/p ovray/pvmpov.html

      For one of my graduate classes I am MPI enabling the latest povray source based on Leon Verrall, Andreas Dilger, & Brad Klines previous work mentioned above.

    4. Re:POVRay by hackstraw · · Score: 1


      POV has been parallelized for years. Believe it or not there is even a sourceforge project for it.

    5. Re:POVRay by AKAImBatman · · Score: 1

      Sorry to reply to myself, but Sun actually suggsted a pretty good use in their FAQ: "Entertainment/Media (digital content creation, animation,rendering,digital asset management)"

      Perhaps this could be used for video encoding? Certainly, many users would probably encode movies they intend to illegally redistribute, but that doesn't mean that there aren't plenty of users who are looking to make a DVD (eventually Blu-Ray) of their home videos or amatuer movies. Sun's grid could potentially compress all the movie pieces in parallel and return the results. A multi-threaded encoder could handle the compression even faster.

    6. Re:POVRay by leoxx · · Score: 1

      I might pay for CPU time if there were a mythtv plugin that used the sun grid to transcode content. Getting many many gigs of data to and from the sun machines might be an issue tho.

    7. Re:POVRay by AKAImBatman · · Score: 1

      POV has been parallelized for years.

      Right you are! I found this version of POVRay that supports DRMAA rendering. So it looks like all that is required is a recompile for Solaris AMD64, and it'll be good to go! :-)

    8. Re:POVRay by drinkypoo · · Score: 1

      It would probably be cheaper to just buy a quad-core system and a copy of cinema craft encoder (assuming DVD, and MPEG2.) As it is, it takes less than 24 hours to make a multi-pass (say, 3 passes) VBR MPEG2 movie at DVD resolution.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    9. Re:POVRay by Wesley+Felter · · Score: 1

      I think the 10GB disk space limit kills that for now.

    10. Re:POVRay by KatTran · · Score: 0, Offtopic

      POVRay has already runs on Solaris 10 on x86-64 hardware.

      http://www.blastwave.org/packages.php/povray

      I'm sure a bunch of other people have binaries too.

    11. Re:POVRay by AKAImBatman · · Score: 1

      POVRay has already runs on Solaris 10 on x86-64 hardware.

      That wasn't my concern. Even if binaries didn't exist for Solaris x64, I'm fairly certain that I could coax the software into compiling. My real concern was whether or not it had been ported to conform to the DRMAA spec. (The standard grid computing API supported by N1.) As it turns out, there are several research versions of POVRay that support DRMAA. So the work is more or less done.

    12. Re:POVRay by Anonymous Coward · · Score: 0

      Perhaps Microsoft can finally release Vista, and we'll all rent the supercomputer time necessary to get acceptable performance from it.

    13. Re:POVRay by dkf · · Score: 2, Funny
      Well, now that you asked, you can use all that computing power to...
      • Compile gentoo with KDE in only 20hrs
      • Browse 10 pages in Firefox
      • Run Windows XP Pro AND Notepad at the same time
      • Get 20FPS in BF2
      • Run a "Hello World" java applet
      Actually, only the compilation stands a chance of working as it is the only one that can work well as a batch job. The others require some kind of interactive display hardware in there, and you can bet you won't have that in the public offering. (Maybe you can get it from the top-end stuff, but if you're big enough for that, you're big enough to have your own supercomputer). My main concern relates to aspects of system security though; their FAQ is seriously deficient on details...
      --
      "Little does he know, but there is no 'I' in 'Idiot'!"
    14. Re:POVRay by Anonymous Coward · · Score: 0

      POVRay runs on the SunGrid and was one of the example applications during its early beta days.

  3. FYI: Non USians need not apply. by Angostura · · Score: 5, Funny

    We're not having you modelling your nukes on our servers thankyouverymuch.

    1. Re:FYI: Non USians need not apply. by MaineCoon · · Score: 4, Funny

      US citizens, however, are free to model nuclear weapons to their hearts' desire. Until Homeland Security shows up.

      --
      Hunt your preferred prey at Aliens vs Predator MUD. Join the war at avpmud.com port 4000
    2. Re:FYI: Non USians need not apply. by spagetti_code · · Score: 1

      unless, of course, you go through anonymizer.com :-)

    3. Re:FYI: Non USians need not apply. by Anonymous Coward · · Score: 1, Funny
      US citizens, however, are free to model nuclear weapons to their hearts' desire. Until Homeland Security shows up.


      Second amendment, baby.

      Let 'em come. :)
    4. Re:FYI: Non USians need not apply. by mwilli · · Score: 1
      US citizens, however, are free to model nuclear weapons to their hearts' desire. Until Homeland Security shows up.

      Should give you plenty of time then!
      --
      My sig beat up your sig.
    5. Re:FYI: Non USians need not apply. by Beltonius · · Score: 1

      Actually, in the Solidworks eDrawings EULA, it states you are not allowed to use the software to develop nuclear, or other weapons of mass destruction. I wonder if Sun's said something similar. http://www.solidworks.com/pages/programs/eDrawings /e2_license.html

  4. Details please by Umbral+Blot · · Score: 3, Interesting

    The ability to but powerful computing time is a cool idea that has been featured in several sci-fi novels. However the article fails to mention exactly how powerful these Sun CPUs are. How much bang do you get for your buck? They also fail to mention how hard it will be to write code for this platform. Can I simply send them some standard C source, or will I have to code using some special extensions that will make my code totally unportable and thus lock me into buying more and more time from them?

    1. Re:Details please by billcopc · · Score: 1

      No code is unportable if it's properly written. Even if they were to use a proprietary API, nothing's preventing you from writing proper abstraction classes (as any good programmer should), that can then easily be retargeted to a different clustering solution with minimal hassle.

      --
      -Billco, Fnarg.com
    2. Re:Details please by temojen · · Score: 1

      And also, how do you get the data to them?

      At work, we find couriering HDDs to have awesome bandwidth.

    3. Re:Details please by The+Snowman · · Score: 1

      TFA and Sun's site are low on details, but I imagine they run Solaris. So if you compile for that OS (which is binary compatible between recent versions) or, more likely, usa Java, you should be fine.

      --
      24 beers in a case, 24 hours in a day. Coincidence? I think not!
    4. Re:Details please by AKAImBatman · · Score: 2, Informative

      However the article fails to mention exactly how powerful these Sun CPUs are. How much bang do you get for your buck?

      Sun claims that they are "dual-core Sun Fire, Opertron servers". That means that they are likely to be something like the V20z which range from 2.0 GHz to 2.2 GHz. It would be nice if they were a bit more specific (e.g. how do you know they'll upgrade the grid in the future?), but their FAQ makes it sound like they're relying on Solaris CPU stats to charge you. OS stats like that are usually based on time slices rather than actual computational power, thus making a "standardized" CPU/hour difficult to create.

    5. Re:Details please by thanasakis · · Score: 1

      Basically you can run any application that runs on Solaris. You will also have to use the Sun N1 Grid Engine to make it run on many processors in parallel. As the article says, there are already companies that are using it to carry out certain tasks. Besides using the Grid Engine software, I suppose you could safely assume a POSIX environment.

    6. Re:Details please by temojen · · Score: 1

      Soooooo not using an interpreted or JiT'd language when you have to pay $1 per CPU hour.... Does gcj work on solaris/Opteron?

    7. Re:Details please by networkBoy · · Score: 1

      you know they wouldn't mind. If you use java their income goes up what 20-30%?
      -nB

      --
      whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
    8. Re:Details please by zwad · · Score: 1

      I don't understand how can it be $1 per CPU per hour? The opteron is only a 2.2 GHz processor how is that a supercomputer?

    9. Re:Details please by networkBoy · · Score: 1

      So with that in mind:
      does anyone have a version of VOBSUB that runs on Solaris 10 (and does it scale beyond 2 cores well)?
      I can see it now: upload a DVD in an encrypted volume as your dataset, have a decryption/encryption wrapper around vobsub and output an encrypted MP4 as your results file.
      Batch up your movies and away you go. Sun will even handle the distrobution of the MP4 as you simply place the download path and password on the net :-)

      Betcha it happens (sooner or later) ;) ;)
      -nB

      --
      whois gawk date unzip strip find touch finger mount join nice man top fsck grep eject more yes exit umount sleep dump
    10. Re:Details please by W2k · · Score: 2, Informative

      Supercomputers aren't about single ultra-powerful CPU:s. A supercomputer consists of lots of CPU:s, possibly thousands, working together. One Opteron 2.2GHz isn't a supercomputer, but a thousand such CPU:s certainly are, if made to work in parallel. Obviously this requires pretty advanced hardware to manage the interconnects and such, in addition to software specifically written for such systems, but that's why everyone doesn't has a supercomputer in his home, even though one can be built using mostly "off-the-shelf" PC hardware.

      --
      Quality, performance, value; you get only two, and you don't always get to pick.
    11. Re:Details please by fritsd · · Score: 1
      I guess the idea is to use a lot of them, like so:
      http://www.top500.org/lists/2005/11/basic (see column 3).
      Maybe like entry #141, the University of Nottingham.

      (I don't want to be a fanboi but) this kind of shows that while people say "Linux is not ready for the desktop" it is ready for certain other tasks that you need computers for:
      http://www.top500.org/lists/2005/11/l/Operating_Sy stem_Family.

      --
      To be, or not to be: isn't that quite logical, Slashdot Beta?
    12. Re:Details please by Anonymous Coward · · Score: 0

      This dumb test compile with 'gcc -O2' indicates that Java is faster.

      fred:~ > count=100000000 sh -c './bench $count && java -server Bench $count'
      time: 2840 ms count=100000000 d=1
      time: 1993 ms count=100000000 d=1.0
      fred:~ > cat bench.c Bench.java

      #include <sys/time.h>
      #include <stdio.h>
      #include <stdlib.h>

      long long ms(struct timeval t) {
        return t.tv_usec/1000 + t.tv_sec * (long long)1000;
      }
      int main(int argc, char *argv[]) {
        struct timeval start, end;
        gettimeofday(&start, 0);
        int i, m = atoi(argv[1]);
        double d = 1;
        for (i = 1; i < m; i++) {
          d += d / i;
        }
        gettimeofday(&end, 0);

        fprintf(stderr, "time: %lld ms count=%d d=%lg\n", ms(end) - ms(start), m, d/m);
        return 0;
      }

      public class Bench {
        static public void main(String [] args) {
          long start = System.currentTimeMillis();
          int i, m = Integer.parseInt(args[0]);
          double d = 1;
          for (i = 1; i < m; i++) {
            d += d / i;
          }
          System.err.println("time: "+(System.currentTimeMillis() - start) + " ms count="+m+" d="+d/m);
        }
      }

    13. Re:Details please by zwad · · Score: 1

      that makes sense, but I still dont get why it is $1 per CPU?? are you saying the expensive part is the special interconnects? Because if you wanted to run a supercomputer of 1000's CPU's for an hour it would cost $1000, that seems really expensive. I mean if I bought one opteron for say $800 it would only take me a year to get 360*24 CPU-hours which at a dollar a piece is way more expensive than just buy the hardware to begin with.

    14. Re:Details please by Anonymous Coward · · Score: 0

      When I change the "benchmark" operation to d += 1 + d/i, so that we calc a number other than '1':
          fred:~ > count=100000000 sh -c './bench $count && java -server Bench $count'
          time: 3143 ms count=100000000 d=18.9979
          time: 2150 ms count=100000000 d=18.997896413850405
      which still puts Java ahead of C.

    15. Re:Details please by dslauson · · Score: 2, Informative
      Here's a link to an FAQ on Sun's site.

      Any code that can be compiled and tested on Solaris 10 can be run on the grid. However, to get the benefit of parallel execution (meaning running parts of a job on multiple processors at the same time), which is really the main benefit of running on a grid like this, you must either write multi-threaded code, or you must use the MPI library, which is pretty much the standard these days for scientific and parallel computing.

    16. Re:Details please by Wesley+Felter · · Score: 1

      If you RTFM, you upload apps and data in zip files (limited to 100MB each and 10GB total).

    17. Re:Details please by Wesley+Felter · · Score: 1

      Sounds like Sun Grid isn't for you. Consider the economics if you need the results of your job tomorrow, not in a year.

    18. Re:Details please by temojen · · Score: 1

      Indeed, that is a dumb test.

    19. Re:Details please by Kadin2048 · · Score: 1

      I worked for a video production company once, a while back...we did a contract job for a local web development company, and one of the video guys was explaining how we did film editing (which involved having the film transferred to some sort of intermediate editing format, working on that to produce timecoded video, editing that and sending it back for printing):

      Video Guy: "So yeah, they take all the film, and put it on video for us to edit; of course now they send it to us digitally, and we send it back--"
      Web Guy: "Wow, how long does it take to send?"
      Video Guy: "Just to send it? Uh, about 10 minutes, I guess."
      Web Guy: "Wow. What kind of a connection do you guys have?"
      Video Guy: (walks to closet, pulls out FedEx Express box) "A lot of these."

      You can fit a LOT of data in a FedEx Medium Box, if you pack it right...

      --
      "Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
    20. Re:Details please by AKAImBatman · · Score: 1

      Soooooo not using an interpreted or JiT'd language when you have to pay $1 per CPU hour....

      Perhaps not JITted code, but for long running processes, a modern mixed-mode JVM like hotspot can kick the pants off a similar C/C++ program. So using the latest Java VM may actually save you money by executing your code as efficiently as possible. You can probably get pretty close with a static compiler by optimizing specifically for the machines that Sun uses, but it would be hard to beat out a runtime JVM that knows the current execution path of the code.

      Does gcj work on solaris/Opteron?

      It should work. But you'll be running with 32 bit instructions, which will probably slow you down considerably.

    21. Re:Details please by greenegg77 · · Score: 1

      Answers to your questions can be found at http://www.network.com/.

      --
      --- This .sig for sale - $500 OBO.
    22. Re:Details please by Anonymous Coward · · Score: 2, Informative
      Their FAQ says their nodes have two single-core Opterons and 4GB of RAM per cpu. I would guess they're using their own X4100 servers with AMD Opteron 254 (2.8GHz). Don't quote me, I don't work for Sun.

      From their FAQ:

      Q: What parallel environments (pvm, mpi, etc.) are available for use on the Sun Grid Utility Services?

      A: MPICH v1.2.6, an open implementation of the "Message Passing Interface" is the only parallel environment currently supported on the Sun Grid. MPICH is configured to leverage IP-based networking in our configuration, and is available on Sun Grid as an included resource for you to use without additional charge.

      If you're familiar with MPI, this should keep you pretty portable. They recommend you test with Solaris 10/x86_64 and gcc3 or Studio 10.

    23. Re:Details please by zwad · · Score: 1

      I see, so the point is it is for people who rarely need to do really large jobs, but do them very quickly.

      So what sort of company/costumer needs to do really large jobs really quickly but doenst need to do them very often?

    24. Re:Details please by MK_CSGuy · · Score: 3, Informative

      From the FAQ:

       
      18. Q:
      What are the basic minimum technical requirements to run an application on the Sun Grid Compute Utility?
      A:
      The following requirements must be met:
      Application must run on Solaris 10 (x64).
      User must own the application or have proper legal licenses to run applications on the Sun Grid Compute Utility.
      Applications must be scripted to work with N1 Grid Engine software.
      Application must be self-contained, with no dependencies on external libraries or data sets
      Application and data sets total size must be under 10GB.
      The user must upload application and data to the Sun Grid over the Internet via portal the Internet Portal at http://www.network.com/

    25. Re:Details please by dubiousdave · · Score: 2, Insightful

      It's not just a question of frequency of use. In order to have your own cluster to do this sort of work, you have to have a lot of infrastructure that many organizations can't really justify. A large datacenter can cost a fortune every year in direct power use by the computers, plus very heavy cooling requirements for the room. You also have to have one or more people running the cluster. Then, on top of all of that, you have to deal with upgrading and maintaining. It's a huge pain that a small, say, fabless semiconductor design firm, may not want to deal with. They focus on their core business of designing circuits and send those designs off to someone else's cluster to test.

      --
      Thank you. Drive through.
    26. Re:Details please by kwark · · Score: 1

      GCJ (4.0.3) can be a piece of crap when it comes to performance, depending on what you are doing. I found it about a factor 6 slower doing some simple disk IO/compression/hashing.

      What it does do is reduce startup time on repeated use by 30%, but considering that the JVM version of that perticular CLI util only takes 0.37s, saving 0.1s isn't worth it (it's only launched about 10 times per hour).

      YMMV

    27. Re:Details please by KatTran · · Score: 1

      It is a bunch of v20z and x2100 style hardware running the Sun Grid Engine which is an open source grid engine that Sun wrote.
      http://www.sun.com/servers/entry/v20z/specs.jsp
      http://www.sun.com/servers/entry/x2100/specificati ons.jsp
      http://www.sun.com/software/gridware/
      http://gridengine.sunsource.net/

      All your code needs to do is run on Solaris 10 on x86 hardware. It can either be 32 bit or 64 bit.

      This isn't a "platform" like a beowulf cluster, it's just a bunch of loosely coupled machines that have a job scheduler in front of them to dispatch jobs. Any of the jobs you submit will only run on a single machine at a time, and it may take 1 or all of the processors available in that machine.

      Sun isn't doing anything magical here, by the end of the summer I plan to have a grid with Solaris SPARCv9, RHEL 4 x64, Windows 2004 x64, and MacOS PPC64 clients, and it will have a web based front end gui scheduler (just like Sun's), and it will be easy to use (just like Sun's), and it will be able to do accounting back to users at the 1 hour interval (just like Sun's) because I will be using _Sun's_ grid engine software to do it.

      I would also point out that you can use the Sun N1 Grid Engine software for free, and that you only need to pay Sun money when you want support.

    28. Re:Details please by gridengine · · Score: 0

      But you didn't count the power and cooling costs... not to mention that space does cost something!

    29. Re:Details please by Siffy · · Score: 1

      It's completely about frequency of use. The costs to pay someone (in this case Sun) to run your jobs will surpass the costs to build, maintain, and run an in house cluster at some point. Say your department only needs to run a job like TFA is referring to for 1 week straight twice a year on 600 CPUs. That would set you back $201,600 (plus tax) a year. But if your department needed to start running those jobs twice a month it would cost $2,419,200 a year ($4032/CPU/half a year). At some point between there it starts being cost effective to set up your own cluster. You shouldn't even have to do the math to realize that Sun wouldn't be selling their cluster's time at this price if it weren't, since Sun is in business to make money. It's all about how many hours of time you'd have to buy a year.

    30. Re:Details please by Siffy · · Score: 1

      Yes. And OpenOffice.org :). I don't know if this would apply or not, but back in 2000-01 I worked for TVA and one of the subdepartments ran weather modeling simulations 24/7 on a too small cluster of 8 Alphas in a rack down in the server room in the basement. This type of stuff would have helped them since they couldn't get the financial approval to purchase a rack powerful enough to get their stuff done. It was literally always overloaded by about 2 weeks of jobs. It brought down the productivity of that group of employees when they were working on that project. But since they always had various projects going on, like I already said, they couldn't get approved for a new server. So something like this would be great for a department that has on going long term projects that they either only work on sporadically or only need computing time say once a month or Fridays after gathering data in the real world the rest of the time.

    31. Re:Details please by Siffy · · Score: 1

      It's still cheaper to build and run your own if you keep it doing stuff 24/365. Else Sun wouldn't be doing this at all as it wouldn't make them any money. Even if you're the military. Take Blue Gene/L + ASC Purple at a combined reported cost of $290M. Combined they have 141312 processors. That would come out to $1,237,893,120/yr if they had to rent that processing time at $1/hr/CPU. Granted the CPUs aren't exactly the same speed/power, but I think it makes my point. Basically you could build a Nuclear Power Plant dedicated to just running the "computer" and a factory that builds air conditioners with the amount of money saved per year by just building your own in this example.

  5. A bug could be costly by DrDitto · · Score: 5, Insightful

    I use grid computing for simulations. If I were charged for CPU-hours, you can bet I would be more careful about debugging. I've wasted thousands of CPU hours because of bugs, or sloppy configuration, in my simulator generating incorrect results. One bug was an infinite loop that resulted in 100 CPUs spinning for a week before I noticed!

    1. Re:A bug could be costly by temojen · · Score: 1

      Have you considered testing on a smaller dataset/serverset?

    2. Re:A bug could be costly by dekemoose · · Score: 1

      Holy crap am I jealous that you could have 100 CPU's maxxed for a week and not notice.

      And holy crap it's lame that that makes me as jealous as it does.

    3. Re:A bug could be costly by ChrisGilliard · · Score: 1

      One bug was an infinite loop that resulted in 100 CPUs spinning for a week before I noticed!

      Your bill would have been $16,800 for that infinite loop.

      --
      No Sigs!
    4. Re:A bug could be costly by Kadin2048 · · Score: 1

      As long as you're going to pay a lot of money for an Infinite Loop, wouldn't you have preferred to buy this one instead?

      --
      "Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
    5. Re:A bug could be costly by chris_eineke · · Score: 1
      One bug was an infinite loop that resulted in 100 CPUs spinning for a week before I noticed!
      Add a cat and some buttered toast and pretty soon we're talking 'bout a black hole, baby!
      --
      "All you have to do is be fragile and grateful. So stay the underdog." Chuck Palahniuk, Choke
    6. Re:A bug could be costly by Pedrito · · Score: 2, Interesting

      Back in the day, I worked for the World Bank writing software on IBM mainframes. Our department was charged back based on our usage. I worked in the telecommunications division and we, like the IBM division, charged departments for their telephone usage. To do this, we'd load the call logs into SQL and then create the bills from that. Our database contained millions of phone calls over the previous year+ of data.

      During development, I made the mistake of doing an unqualified join between the primary call table and the table of departments (of which there were probably a few hundred. Pretty simple math... The result set was somewhere in the hundreds of millions to billions. I realized the problem the moment I submitted the job. Unfortunately, there was no way for me, a lowly user, to kill the job, once it began. By the time we managed to get the job killed, I had squandered thousands of dollars in computing charges. My boss was none too happy. Ah, the good old days...

    7. Re:A bug could be costly by geekoid · · Score: 1

      can I ahve your resume, just in case.*

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    8. Re:A bug could be costly by timeOday · · Score: 1
      Your bill would have been $16,800 for that infinite loop.
      As it stands, the mistake may have cost his company more or less than that; we don't know (and probably neither do they). The only real question is whether you can provide CPU time cheaper than Sun will sell it to you.

      Personally, I think $1/CPU hour is a reasonable place to start. Maybe they'll let the price float with demand in the future.

    9. Re:A bug could be costly by Heembo · · Score: 1

      One bug was an infinite loop that resulted in 100 CPUs spinning for a week before I noticed!

      How did you have 100 processors maxs out for a week without knowing?

      --
      Horns are really just a broken halo.
    10. Re:A bug could be costly by Diag · · Score: 1

      You didn't have authority to cancel your own job? In a development environment? *shrug* Each shop is different, I suppose.

      The biggest bill I ever saw on a mainframe was when we (Operators) discovered how to display crude graphics to an ISPF terminal, and started playing with Mandelbrot generators for fun. That only lasted a month, until the bill came in.

      --
      Serving Suggestion: Defrost
    11. Re:A bug could be costly by andy_t_roo · · Score: 1

      because you could expect your program to use 1000 cpu's for a few days, and only notice the mistake when you program didn't stop

    12. Re:A bug could be costly by Pedrito · · Score: 1

      You didn't have authority to cancel your own job?

      Of course, I could kill my job, and I did, immediately, but once the SQL side of the job began executing, it was running under the shared SQL account, not mine. That one, I didn't have the authority to kill.

  6. How's this work? by the_humeister · · Score: 2, Interesting

    Is it like getting an account on someone's server and then being able to do whatever the hell compute-intensive work you want? I can't seem to find the relavent details, or my Parkinson disease is kicking in.

    1. Re:How's this work? by the_humeister · · Score: 1

      OK, answered my own question after random clicking. According to this your app needs to be able to run on solaris 10. Then just upload it and it runs.

  7. Obligatory by PitaBred · · Score: 4, Funny

    Imagine a beowulf cluster of... oh, forget it

    1. Re:Obligatory by Douglas+Simmons · · Score: 1

      Somebody already said it in a much funnier way than you did. And without that lame ass parenthetical apology or using AC.

  8. Since I don't remember the earlier stories... by TubeSteak · · Score: 1

    Do you (the customer) supply the software to run on these distributed boxen?

    Cause if that's the case, I can see a business model that involves lophtcrack or John the Ripper.

    --
    [Fuck Beta]
    o0t!
    1. Re:Since I don't remember the earlier stories... by AKAImBatman · · Score: 2, Insightful

      Do you (the customer) supply the software to run on these distributed boxen?

      Yes. See the Sun FAQ.

      Cause if that's the case, I can see a business model that involves lophtcrack or John the Ripper.

      Considering that Sun has your Credit Card number and your contact information, I highly doubt you'll be getting away with much DDOSing or Spamming. Even if you use a stolen card #, Sun would be likely to kick you off the servers after they notice excessive network and/or mail server usage.

    2. Re:Since I don't remember the earlier stories... by Stradenko · · Score: 1

      I fail to see the connection between a pair of password guessing/cracking utilities and spam/DDOS....

    3. Re:Since I don't remember the earlier stories... by AKAImBatman · · Score: 1

      My mistake. I thought he was making up script kiddie names. A quick Google search has disavowed me of that notion. Of course, you still need to harvest those passwords from somewhere...

    4. Re:Since I don't remember the earlier stories... by Anonymous Coward · · Score: 0

      They'd be trying to find the passwords for the systems making up the Sun grid ;)

      Though somehow I suspect each account will be placed in a virtual environment/jailed, with random passwords, and probably disappear at set time after use. If not, they should be.

    5. Re:Since I don't remember the earlier stories... by Stradenko · · Score: 1

      What would be interesting is using the grid to generate rainbow tables...generating rainbow tables and selling them on DVD is a lucrative business. (well, except for LANMAN tables -- which shmoo gives away gratis)

    6. Re:Since I don't remember the earlier stories... by Slithe · · Score: 1

      I do not think that L0phtcrack runs on non-Windows platforms. John the Ripper does, though.

      --
      ---- "XML is like violence. If it doesn't fix the problem, you aren't using enough."
    7. Re:Since I don't remember the earlier stories... by TubeSteak · · Score: 1

      I agree wholeheartedly with the sentiment, but a lot of crypts use a salt, which basically makes anything but brute-forcing worthless.

      http://en.wikipedia.org/wiki/Salt_(cryptography)

      Of course, if the salt is available to you, then making a rainbow table using a grid is much more efficient than trying to brute the system

      --
      [Fuck Beta]
      o0t!
  9. Finally.... by svunt · · Score: 1

    I can outsource physics rendering at LAN parties and show my friends who's 1337 ;)

  10. Before anyone shouts :DUPE! by scenestar · · Score: 2, Informative

    Casual Sun observers will be scratching their heads right about now, believing that Sun had already announced such a service a long time ago. That's correct.

    rtfa kthnx

    --
    perpetually dwelling in the -1 pits
  11. Free demo here... by GillBates0 · · Score: 3, Interesting

    Click here to kick off a job on Sun's Compute Grid consisting of AMD Opteron-based Sun Hardware.

    --
    An Indian-American Hindu committed to non-violent thought/speech/action alarmed by the global explosion of radical Islam
  12. Maybe it's like razors by PIPBoy3000 · · Score: 1, Flamebait

    The CPU is cheap, but you'll be paying an arm and a leg for "extras" like disk storage and memory.

    Want a printout of your results? That's $100 per page . . .

    1. Re:Maybe it's like razors by PitaBred · · Score: 1

      But the CPU being cheap is the point. Not everyone wants to set up their own cluster, but lots of people could use a cluster periodically, and would do so more often if it was relatively cheap. Sun is basically catering to them.
      I mean, a small-ish game developer could pre-compute a lot of data on the grid without having to invest in their own hardware and expertise and time for a less-capable solution, all kinds of things really.

    2. Re:Maybe it's like razors by TubeSteak · · Score: 1

      In the FAQ
      http://www.sun.com/service/sungrid/faq.xml#q24

      They mention that (for a limited time) you get a max of 10GB of storage (for 180 days!) and that each box has 4GB per CPU.

      So... maybe in the future, you'll pay extra for disk storage, but RAM will always be 4GB per CPU

      --
      [Fuck Beta]
      o0t!
    3. Re:Maybe it's like razors by Anonymous Coward · · Score: 1, Informative

      Why is this false accusation modded up?

      Read the FAQ.
        - The $1/CPU-hour is the only charge.
        - 10GB of disk storage is available per account.
        - 4GB of RAM is installed per AMD64 CPU.
        - No print services are for sale.

      Also,
        - No upload/download fees.
        - MPI is used for distributed jobs. Compile with gcc or the now free Sun compilers.
        - Grants to schools are available.
        - Developers can get 100 free hours.

    4. Re:Maybe it's like razors by Anonymous Coward · · Score: 0

      And would you care to offer how you came up with that conclusion? Sun does not say anything about charging extra for memory or diskspace.
      Nothing about printouts either. Rather, the model is that you upload your app, run it and get the results back as a file(mp3's, text, xml whatever). Download the file and you are done.
      Harsh Chaudhary.

  13. Imagine All the SPAM! by siefkencp · · Score: 1

    Just think.

    Upload your app, pay 1$ a CPU hour, charge to a credit card with a stolen identity...

    Boom! -- Enter the biggest mail bomb credit card theft combo in history...

    1. Re:Imagine All the SPAM! by Wesley+Felter · · Score: 2, Insightful

      That's why the grid nodes are not connected to the Internet.

    2. Re:Imagine All the SPAM! by Anonymous Coward · · Score: 0

      no. upload cracker, use irregular cc account.
      boom. bigger fun.

  14. Greetings Professor Falken... by boldtbanan · · Score: 5, Funny

    I wonder if I can play games on Sun's system. Perhaps a nice game of chess? Or maybe Global Thermonuclear War?

    1. Re:Greetings Professor Falken... by professorfalcon · · Score: 1

      Don't forget Tic-Tac-Toe. On 5,000 processors.

  15. Sun Grid HW / SW specs by DAldredge · · Score: 4, Informative

    From their FAQ:

    Q:
    What are the components of the Sun Grid Compute Utility?
    A:
    The Sun Grid Compute Utility service consists of the following parts:

            * Sun Fire dual processor Opteron-based servers with 4GB/RAM per CPU
            * Solaris 10 (x64)
            * Solaris 10 OS;
            * Sun N1 Grid Engine 6 software;
            * Grid Network Infrastructure of 1Gb switched Data Network and 100 Mb dedicated management network;
            * Web-based access portal; and
            * Internet-only access to upload data and applications (no physical access to location);
            * Storage allocation of up to 10 GB per user account.

    http://www.sun.com/service/sungrid/faq.xml

    1. Re:Sun Grid HW / SW specs by Anonymous Coward · · Score: 0

      They also claim the 10GB of storage is free for a limited time, all you need to do is have some account activity every 180 days. Run one $1 job every few months, backup 10gb offsite.

    2. Re:Sun Grid HW / SW specs by pimpimpim · · Score: 1
      Their network seems to be a pretty basic 1Gb network. If you want to do parallel jobs that need really fast communication, this will not be suitable for you.

      Let them put infiniband on those systems and it becomes interesting. In the meantime I'd say it's nice for single CPU jobs only.

      --
      molmod.com - computing tips from a molecular modeling
    3. Re:Sun Grid HW / SW specs by Trinn · · Score: 1

      Actually if your job can be parted out into parallel batches that only have to sync every so often, rather than being a heavily multithreaded job, it will probably run fine on that system. Basically its just a design consideration, you have to take into account the limited bandwidth when coding your app, and emphasize doing work in chunks, and probably staggering the chunks so that the control app can aggregate the results without waiting

  16. What, they are trying this again? by Zangief · · Score: 2

    Wasn't their previous attempt to rent CPUs a failure?

    I remember an article in slashdot about how the Sun grid was completely unused.

    1. Re:What, they are trying this again? by Anonymous Coward · · Score: 1, Interesting

      Not completely unused, it has been up and running for over 6 months to selected customers who have been running Proof of Concepts and performing Engineering Development work.

      I think Sun certainly started shouting about it too early, but this is the same service that was announced way back - Sun has finally got the technology/security/usability/stability/etc sorted out and gone live.

    2. Re:What, they are trying this again? by dotpavan · · Score: 1
  17. The Sun is setting by hackstraw · · Score: 3, Interesting


    Wow, $1/CPU/hr. Same price as an MP3 off of iTunes, so it must be worthwhile, right?

    OK, we are only about 3.5 months into the year of 2006, and lets look at some real data:

    I run a few small to medium sized HPC clusters, and on one of them, here are the CPU hours used during 2006 -- 163,000+ this is on less than $500k of hardware that is years old. That would cost $163k just in computing time, not to include time to port applications, debug, etc.

    Sun needs to be run by engineers and visionaries again, not by marketers. $1/CPU/hr is not going to do much better on those falling stock prices than selling $200 Linux PCs in Wal-Mart.

    1. Re:The Sun is setting by Anonymous Coward · · Score: 0

      That would cost $163k just in computing time, not to include time to port applications, debug, etc.

      The thing is, you're *only* paying for time. You're not paying for a huge amount of hardware up front, or the space to keep it in, or the power it consumes, or the people to keep it running, or any number of other expenses. (Well, you are paying for that indirectly, but the point is that it's abstracted behind the $1/CPU/hr. price.)

      Presumably this comes out cheaper in many scenarios.

    2. Re:The Sun is setting by Anonymous Coward · · Score: 0

      Hmmm. Last I checked we are only 2.5 months into the new year

    3. Re:The Sun is setting by eclectus · · Score: 1

      Oh, please. So a company releases something that you don't need, and you want to complain about it? A couple of questions for you? Your 163,000 hours of compute time on old hardware -- how much of that time was your cpu actually in use? Were you were your CPU's pegged for the whole year? Second, they are using brand new gear, not 'years old' equipment, so that ought to cut down on the number of hours of computing.

      Maybe this offering isn't of use to you. However, if you have bill runs that takes 16 hours each day to run on your current gear, and you can farm it off to Sun and get the data back in an hour, that might be pretty handy for a business.

      --
      This signature is a waste of 42 characters
    4. Re:The Sun is setting by rsclient · · Score: 1

      Maybe there's more to your figures than you're letting on, but frankly you're not selling your point very well.

      Just sitting in your office, you cost your company about $150,000/year. They could lay you off, and use that money to pay Sun. They can then make up the difference by selling off their computers, netting, say, $100,000. They stop writing huge checks to the local power company, and sub-lease your computer space. They also save money on hardware fixes for the computers in your cluster, and on the cost of technicians.

      From what you've said, Sun looks cheaper than your current solution (except for the porting, but heavens, that's not very expensive these days!)

      --
      Want a sig like mine? Join ACM's SigSig today!
    5. Re:The Sun is setting by Anonymous Coward · · Score: 1, Insightful

      Cost of writing applications and porting them should be included in any estimate, yet you only include it in the estimate of Sun's cost and not your own. Also worth pointing out is that your "years old" hardware wont perform anywhere near the level at which suns offering will.

      You also failed to mention anything about this "less than $500k" cluster. For the price of a $499k cluster (which again doesnt include man hours spent on code, so it wont be counted against sun either) I could rent 50 cpu's 24x7x365 x1. You get more cpu power buying a system at that price point. Now factor in maitenance costs. Failure costs. Power and cooling costs. Personell costs for admining the setup (not coding). You still get the edge buying the system, but by how much ? The reality is this is not geared towards computer people. Its geared towards math, graphics, and physics people. Groups who dont always have the time and resources to buy, setup and admin a cluster but who have some data once in a while that they need to proccess.

      Now apply your math to someone who has something that needs to be done once in a while in a condensed time frame. This is one hell of a lot better system for that then buying a cluster and hoping it doesnt break right when you need it ... or while it sits idling for the majority of the year.

    6. Re:The Sun is setting by __aaahtg7394 · · Score: 2, Insightful

      It gets even better when you consider that Sun's smaller Opteron kit starts at about 2k$/node. So, if you need more than a few months' processing, you can just buy the boxes and build the infrastructure for about the same cost.

      I suspect the real selling points are:
      - Sun's service is probably straightforward for app developers.
      - The hardware is essentially "infinite."
      - "Oh, you need a month's worth of processing done by next Monday? We'll have it done Saturday night, if you'd like to pick it up then, ma'am."
      - Two words: volume pricing.

      But, I have a hunch that the real point is marketing. "Oh, you need a 1,000 node grid? We happen to have expertise in building just such a thing. Here, give this information to your apps people, let them try it out, see if they like it. We can install just this setup for you here, if you'd like."

    7. Re:The Sun is setting by Tsunayoshi · · Score: 2, Insightful

      The whole point of this was stated by John Schwartz when the idea first came up. This is moslty for the market of people who only do massive number crunching on a few occasions but often enough that they would maintain their own computing cluster. Now, these same people can get rid of the hardware, people, power, cooling, and lab space and the other costs associated with maintaing a cluster and just worry about paying for CPU time the few occasions they actually need it. And if suddenly you have an immediate need for 10x the computing power, let Sun know and you can get more systems involved for your project. If you maintained your own compute cluster, how easily could you add 50 cpus to a task if you are already running at 100% CPUS??

      --
      "Get a bicycle. You will not regret it, if you live." - Mark Twain, "Taming the Bicycle"
  18. Doesn't the pricing encourage non-CPU intensive? by wsanders · · Score: 1

    I mean, hell, here's 10TB of data that I'm currently backing up to tape.

    Do you want me to package it as a J2EE WAR file? Fine!

    --
    Give a man a fish and you have fed him for today. Teach a man to fish, and he'll say "WHERE'S MY FISH, YOU IDIOT?"
  19. I would have made use of Sun's grid already by c0l0 · · Score: 2, Interesting

    if it boasted a 64bit Java VM. A mate of mine does some very interesting research in number theory, and a few of his applications would need massive amounts of fast addressable memory. 64bit of address space would conveniently suffice, i suppose. Any suggestions on what else (cheap, or at least affordable) to consider using, anyone?

    --
    :%s/Open Source/Free Software/g

    YTARY!
    1. Re:I would have made use of Sun's grid already by shaitand · · Score: 1, Insightful

      You could save some really serious money by porting to a proper compiled language.

    2. Re:I would have made use of Sun's grid already by drix · · Score: 0

      Just curious, what sort of number-theoretic problems are you tackling that would warrant using Java?! Talk about a memory hog. Most of the programs of this kind that I have seen are computationally intensive but easily expressed in 100 lines of C. In fact, most of the oustanding problems themselves can be written in one or two sentences (the Riemann hypothesis comes to mind).

      --

      I think there is a world market for maybe five personal web logs.
    3. Re:I would have made use of Sun's grid already by Anonymous Coward · · Score: 0

      Are you sure, in fact, that a 64-bit JVM isn't part of the grid?

      Solaris ships with a 64-bit JVM for both their x64 and sparc systems.

    4. Re:I would have made use of Sun's grid already by anagama · · Score: 1

      Probably not "cheap" by any stretch of the imagination. http://www.sgi.com/products/visualization/prism/st reaming_video.html

      Scroll down to the GeoProbe movie -- they mention a system with 512 GB of "shared memory". And mention linux.

      --
      What changed under Obama? Nothing Good
    5. Re:I would have made use of Sun's grid already by Richard+W.M.+Jones · · Score: 1
      If he's writing it in Java, no wonder he needs a supercomputer to run it.

      Rich.

    6. Re:I would have made use of Sun's grid already by kamg · · Score: 2, Informative

      Sun's JDK has 64-bit support for both sparc(v9) and amd64. Just run with the '-d64' flag.

  20. Let's get these out of the way by Anonymous Coward · · Score: 0

    In Soviet Russia, grid process runs you! Immagine a beowolf cluster.... etc...

    1. Re:Let's get these out of the way by PitaBred · · Score: 1

      In Soviet Russia, 'imagine' spells itself correctly (sorry, couldn't resist. Damn lack of impulse control...)

  21. Ok, so by ch-chuck · · Score: 3, Funny

    where do I submit my deck of fortran punch-cards and where do I pickup the printout?

    --
    try { do() || do_not(); } catch (JediException err) { yoda(err); }
  22. No customers after three years... by Anonymous Coward · · Score: 1

    Sun has been offering this service for almost three years now, and to my knowledge they still have had zero customers. IBM seems to be the only utility computing supplier with any customers using their BlueGene solution in Yorktown. Virginia Tech even offered cycles to outside parties for $0.40 CPU/hour and had no takers. It's gonna be a while before something like this gets any traction. People are way too protective of their data to allow it off-site.

  23. Solaris-10 or Java Binaries only? by Anonymous Coward · · Score: 0

    Sorry - but its been well over a decade since I touched Solaris - maybe if they'd allow Linux binaries.

    1. Re:Solaris-10 or Java Binaries only? by mondrian · · Score: 2, Informative

      You can run unmodified Linux binaries on Solaris 10 thanks to Janus.

  24. video encoding by TubeSteak · · Score: 1

    The time it'd take to send/upload the content would probably be longer than if you just plunked down and did it yourself on any relatively new computer.

    The only way it'd make sense is if you have a lot of video to compress and even then, over the long term it might still make sense to buy a cheapo computer or two & DIY.

    --
    [Fuck Beta]
    o0t!
  25. Isnt this really expensive? by zwad · · Score: 0

    I don't understand how can it be $1 per CPU per hour? The opteron is only a 2.2 GHz processor how is that a supercomputer?

    1. Re:Isnt this really expensive? by Kadin2048 · · Score: 4, Interesting

      That's 2.2 GHz per processor, times about a thousand processors or so. That's how modern supercomputers work. The processing nodes themselves are somewhat unimpressive, but they're built so that they scale really well, and deal with problems that are designed so as to be broken up into lots of little parts and solved simultaneously. So if you used all the processors on the machine for an hour, your bill (theoretically) would be $1,000.

      The most powerful computer in the world right now, ASC Purple (it does nuclear weapons simulations for the USG), has 1.5 GHz RISC processors. Not exactly impressive, by today's standards ... except that it has something like 12,000 of them.

      It's the infrastructure to get that many processors (and their associated dangly bits) talking to each other and working on the same problem efficiently that's expensive and nontrivial.

      --
      "Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
    2. Re:Isnt this really expensive? by zwad · · Score: 1

      I see, so the point is it is for people who rarely need to do really large jobs, but do them very quickly.

      So what sort of company/costumer needs to do really large jobs really quickly but doenst need to do them very often?

    3. Re:Isnt this really expensive? by sokoban · · Score: 1

      ASC Purple 63.39 TFlop/s 10240 processors

      Blue Gene W 91.29 TFlop/s 40960 Processors

      Blue Gene/L 280.6 TFlop/s 131072 Processors

      --
      09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0 is the magic number.
    4. Re:Isnt this really expensive? by Jeng · · Score: 1

      What he's saying is still correct though.

      WTF? $1 to run a CPU for 1 hour

      It being 100 CPUs at $100 an hour makes no difference.

      Now if it was $1 to run on 100 CPU's for 1 hour, well, now we're talking.

      As is it costs me a couple pennies a day to run an app on my 2.2ghz athlon64.

      --
      Don't know something? Look it up. Still don't know? Then ask.
    5. Re:Isnt this really expensive? by patio11 · · Score: 1

      One of the target markets for the Gifu Grid (a prefectural initiative here in Japan which we're supporting at my place of employment) is architectural firms doing earthquake simulations. I'm not privvy to the specifics, but apparently their big use case is spend a lot of man hours making changes to the building plan, get to Friday, hit "test for structural soundness" and go out drinking. When you're on a n-week construction cycle, spending two weeks waiting for the results of your compile (oops, sorry, at a 6.2 quake with certain assumptions this fails... BACK TO THE DRAWING BOARD!) costs you both money and your completition-on-time goal. But during the vast majority of the project cycle you *don't* need a server farm -- you've got X desktops running your favorite CAD software, which is all well and good, but if you had a supercomputer or beowulf cluster in the back room it would be powered down 98% of the time. So, theoretically, we could sell them time on our network -- we'll have your simulation done by Monday, you don't worry your architectural heads about setting up a Beowulf cluster.

    6. Re:Isnt this really expensive? by patio11 · · Score: 1
      When you're looking at an Estimated Time To Solution measured in years (for a trivially codable problem with no real-world utility which we used to stress-test our grid, try n-queens with N at 21 or 22), you start to realize that a) pennies per hour does not scale and b) depending on what you're running, you'd think paying $1 per CPU hour was cheap compared to the eventual output you get.

      Here's an example: when your $1 billion (thats billion with a B) commercial highrise in Japan is undergoing earthquake testing, "Screw the grid alternative, I've got an Alienware at home we can run this on, should have the results back by 2127 or so..." and paying for, oh, 100,000 CPU hours (pulling number of of thin air) looks like a very attractive proposition (consider what you'd have to spend as an architectural firm to afford your own supercomputer or grid to make that happen, plus the maintenance and employee costs to keep it running, and the fact that you get no use of it for weeks or months at a time in between earthquake tests but the tests need to be done right and done fast so you can get on with construction).

    7. Re:Isnt this really expensive? by hpcanswers · · Score: 1

      According to http://www.top500.org/, ASC Purple is number 3. Number 1 is BlueGene/L and number 2 is BlueGene W.

  26. This sounds familar by Matt+Perry · · Score: 2, Interesting

    I thought that Sun already had their grid available and that no one wanted to use it because they would have to agree to be in a marketing campaign. Is this still the case? The terms of service on the network.com site redirects to an error page.

    --
    Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
  27. Real TIme Fractal Exploring by Anonymous Coward · · Score: 1, Funny

    This could be used to interactively zoom in and out of interesting fractal patterns with a high refresh rate.

  28. Fast Gentoo install by slicersnatch · · Score: 1

    Does this mean that we can now all get gentoo up and running in less then 50 hours?

  29. Sure way to boost Sun's profits... by mythz · · Score: 1

    Fire the current CEO and marketdroids. Apart from saving the salaries of a higly paid and underperforming execs, companies might actually want to do business with a company who's chief isn't soley concentrated on funneling as much money out of you as they possibly can. Thats if there are still some talented engineers who haven't deflected yet.

    Scott McNealy they need you over at SCO.

  30. At last a solution for h264 DVD recoding!! by OlivierB · · Score: 3, Interesting

    If I could compile mencode/mplayer for Solaris I could upload my dvd isos and get sun to encode these for me in H264 for my HTPC.
    I anticipate that each film would cost me ~$2. Not bad. Is that a safe bet? ANybody know what disk space they give for "personal files".
    Now to explain to my ISP that I am not participating in illegal file sharing with +100GB per month of traffic is not going to be easy..

    More seriously, I could use this to run some of my Monte-Carlo simulators..

    --
    Artificial intelligence is no match for natural stupidity
    1. Re:At last a solution for h264 DVD recoding!! by Slithe · · Score: 2, Informative

      The quota is 10 GB per account. Each CPU has 4GB of associated RAM. Is that enough space to encode a DVD?

      --
      ---- "XML is like violence. If it doesn't fix the problem, you aren't using enough."
    2. Re:At last a solution for h264 DVD recoding!! by dodobh · · Score: 1

      Upload DVD. Encode. Download results. Remove old data. Loop.

      --
      I can throw myself at the ground, and miss.
    3. Re:At last a solution for h264 DVD recoding!! by Anonymous Coward · · Score: 0

      mplayer is available for download at Blastwave and you could try asking for mencode.
      Check SunFreeware for precompiled packages as well.

      If you want to build things yourself, Solaris 10 and the Sun compilers are free for download and use. gcc is also included with the OS but is slower.

      You can find OS source code at OpenSolaris.

  31. Price in Dollars for hours? by Mantees+de+Tara · · Score: 1

    $1 per CPU per hour

    On Slashdot the price should be published in WoW gold per SETI units elaborated.

  32. 56K modem line, of course! by drgonzo59 · · Score: 1
    You send your 100GB of data over a 56K modem line.... After a couple of weeks your data will get there, then it will finish running super fffffast, (no, really!), and in about another couple of weeks you'll get it back.

    I think they should just add a service where people could deliver or snail mail dual layer dvds or tapes with data, with a set of new dvds or tapes for results. One can overnight it with Fedex. Then, they run the program on the grid, load up the result on another set of dvds or tapes and overnight them back to you.

    1. Re:56K modem line, of course! by stm2 · · Score: 1
      I think they should just add a service where people could deliver or snail mail dual layer dvds or tapes with data, with a set of new dvds or tapes for results. One can overnight it with Fedex. Then, they run the program on the grid, load up the result on another set of dvds or tapes and overnight them back to you.


      If you change "DVD" and "tapes" for "punch card" in your sentence, you are describing computing as it was known 30 years ago.

      --
      DNA in your Linux: DNALinux
    2. Re:56K modem line, of course! by drgonzo59 · · Score: 1

      Nothing new under the Sun (pun intended) I guess...

  33. Is there a distributed alternative? by gfody · · Score: 3, Interesting

    Like seti or folding@home except instead of donating your spare cpu cycles for one particular task you'd be making them available for anyone to rent?

    The price per hour per cpu could be based on demand and could be distributed to all the contributers. Imagine all the processing power out there not being used. Especially the gpus on people's video cards while they're not playing games.

    --

    bite my glorious golden ass.
    1. Re:Is there a distributed alternative? by Tiger4 · · Score: 2, Funny

      "except instead of donating your spare cpu cycles for one particular task you'd be making them available for anyone to rent?"

      Yes this exists, but I think the 'Bot-ware manufacturers have the market sewn up, and at $0 / hr.

      --
      Behold, this dreamer cometh. Come now, and let us slay him... and we shall see what will become of his dreams.
    2. Re:Is there a distributed alternative? by Wesley+Felter · · Score: 1

      A few companies tried this about five years ago, but it turns out the idle time on a random PC is worth less than the overhead cost of using it.

    3. Re:Is there a distributed alternative? by 4D6963 · · Score: 1
      Or better, have a F@H type utility running other people's computations, thus earning "credit" and then spend your credit on your own computations when you need.

      In other words it would mean saving your own CPU time to use it later. I used to dream about this back when I had a 75 MHz PowerPC and was doing 3D in raytracing on it.

      --
      You just got troll'd!
  34. What types of useful applications? by sumdeus · · Score: 1

    So could someone explain to me what types of applications and organization would most benefit from this? Most larger scale operations already have an internal cluster setup, so that would eliminate them from contention. Since there is not web based API at this point, and not web interaction I can't imagine it being used for any time of spamming of any type. So the question remains, where is the business need?

    --
    Peter: I got an idea, an idea so smart my head would explode if I even began to know what I was talking about.
    1. Re:What types of useful applications? by geekoid · · Score: 1

      older mainframe systems where the number of jobs keeps increasing, but don't have the immediate resourse for an upgrade.

      Also, jobs that run once a year might be cheaper to run this way, as opposed to new hardware that won't be used through most of the year.

      Is a dollar an hout cost effective? I ahve my doubts, considering most of the cost is the support for the application.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
  35. CPUShare by PenGun · · Score: 2, Interesting

    Oh please. We have cpu time for 10c an hour over at Andrea Arcangeli's CPUShare website:

    http://www.cpushare.com/

      Still experimental for now but soon ...

      I did some math and I will build a server farm if I can get a steady 10c/hr/processor.

          PenGun
        Do What Now

    1. Re:CPUShare by Nugget · · Score: 1

      Sounds like maybe "have" is a bit optimistic. From the site it looks like they may have a competing network at some undefined point in the future if they ever get around to building it in their spare time.

    2. Re:CPUShare by PenGun · · Score: 1

      Well we have our seller interface. We are experimental for now but I expect that to change when the patents are in place.

          PenGun
        Do What Now ??? ... Standards and Practices !

  36. Condor by Isaac-Lew · · Score: 2, Insightful

    Why wouldn't a mid- to large-sized organization use something like Condor? Just install it on everyone's (linux or win2k/xp) server/workstation, maybe set some prioritization scripts so that it would use more resources after-hours (when most people are out of the office, but have their systems on anyway), & save themselves $$ instead of paying to have their data on someone else's remote system?

  37. Seti@home by way2trivial · · Score: 3, Funny

    anyone have any idea how much it would cost me to buy the # one spot on boinc?
    Take down NEZ for one day-- that would be sweet

    http://www.boincstats.com/stats/boinc_user_stats.p hp?pr=bo&st=0&to=100

    --
    every day http://en.wikipedia.org/wiki/Special:Random
  38. Greatly speeds up debugging! by Anonymous Coward · · Score: 0

    I use our 80 Opteron cluster for debugging: send 80 different data sets through the same code and you hit corner cases about 80x faster. Nice.

  39. If only something like this were free by slagell · · Score: 1

    If only something like this were free.

    Oh, wait Xgrid.

    1. Re:If only something like this were free by Heembo · · Score: 1

      Ty'ea right, the moment Apple hardware becomes free please wake me up from my nap.

      --
      Horns are really just a broken halo.
  40. Our cost would be much lower by wgadmin · · Score: 1

    This is interesting. We have a 1600 CPU Linux cluster and we recently did an envelope calculation in regard to how much a single CPU hour would cost (based on initial investment, admin salaries, number of CPUs, projected number of processed jobs, projected end-of-life duration, etc) and we came up with $0.09/hour.

    1. Re:Our cost would be much lower by Anonymous Coward · · Score: 0

      Yeah, but that's a) in bulk and b) you have the cluster all year round. If you only need one for a week of work (which needs to be done that week, so you can't just use fewer CPUs for longer) then the price for getting that work done goes up by 50. Plus, did you include the cost of the building?

      It's not entirely stupid, in fact I'm tempted to give it a 1-proc spin for a day just to run a benchmark and see how it compares to our very modest cluster, in case I max it out and need more processing in a hurry. I'm not a US citizen though :(

  41. Digital Fortress by nneonneo · · Score: 1

    Anybody read Digital Fortress?

    1. Re:Digital Fortress by ArmedStupidity · · Score: 0

      Mmhmm. Good book. Does sound a bit like TRANSLTR, doesn't it? Except that thing was massive and relied on extremely unrealistic amounts of paralell processing inside a single machine, instead of this cluster(f*ck) Grid deal.

  42. how powerful a cpu? by roror · · Score: 1

    Any idea how powerful a cpu I get one hour of for $1?

  43. Are there competitors? by emin · · Score: 1

    Are there any competitors to the Sun Grid?

    For example, is there a similar service for LINUX machines? I could certainly use a lot of extra machines to run simulations, but I mostly use LINUX. Having to switch everything over to Solaris makes this service much less useful.

    Thanks.

    1. Re:Are there competitors? by mondrian · · Score: 1

      You can run Linux applications unchanged on Solaris 10 thanks to Janus.

  44. Christ, you've got some good upload! by Ayanami+Rei · · Score: 1

    I imagine for most residential (and even some commercial) broadband users the latency of transferring the 4GB+ of VOB files outweighs the encoding time even on a modest desktop CPU

    --
    THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
  45. Slippery Slope by Nikker · · Score: 1

    This is all good with todays horsepower but what kind of company is going to upgrade a system like this in the famous 18 month period.

    If Mores law holds true Suns revenue will half evrey 18 months, so the question is, is this permanent? Will they find new things to burn clocks? Or will they try to corner the market and still try to charge $1 according to a 2GHz machine.

    The only way I could see this working is if they charge by clock rather than time and charge very little, this way you pay the same amount in 3 years but you get the result in 1/4 the time.

    --
    A loop, by its nature, continues. If that didn't make sense, start reading this sentence again.
  46. Hit By a Network Attack? by Heembo · · Score: 2, Interesting

    Oh comon, they are just trying to cover-up getting slashdotted!

    New.com article "Sun Grid hit by network attack" : http://news.com.com/2100-7349_3-6052968.html

    --
    Horns are really just a broken halo.
  47. Anyone else think.... by kadathseeker · · Score: 1

    Skyne- no way. Nevermind. It'll never happen.

    Unless I install this neat little proggy I wrote...

    --
    The 'Net is a waste of time, and that's exactly what's right about it. - William Gibson
  48. Sun Grid? by numbware · · Score: 1

    That's nothing. Where's my Google Grid?

    --
    I'm going to go create my own technology news site, with blackjack and hookers. You know what? Forget the news site.
  49. On-Demand Computing? by hpcanswers · · Score: 1

    IBM was pushing computing-as-a-utility a few years ago. Their premise was that just as customers don't have their own power generators—they just buy electricity at market rates exactly as required—folks should be able to acquire computing time only as needed and let a company of professionals deal with all the maintenance issues. But no one bought into on-demand computing and IBM has since backed away from this position.

    The reason this flopped for IBM was that everyone compared it to the "good old days" of dealing with shared computing facilities and batch loaders. It's the same reason people buy small clusters instead of paying for space with a national lab. The comparison to punch cards is apt, as are the discussions of "what if I spend all this time and money only to discover that I bug?"

    I don't see how Sun is going to succeed in the market. Smaller customers will probably be better off buying a "personal supercomputer" rather than rent time. As for the really large customers, they will be running jobs 24/7 for years on lots of processors; it may in fact be cheaper for them to buy a BlueGene than rent time from Sun.

  50. Thanks. by Kadin2048 · · Score: 1

    I stand corrected -- you're right, it's now number 3.

    There was a time when it was number 1, but damned if the folks at IBM haven't been busy in the meantime.

    Computers...don't pay attention for a while, and suddenly everything you know is wrong. :)

    --
    "Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
  51. Meanwhile, a grid that's actually doing paid work by Animats · · Score: 1
    Unlike the Sun Grid, which is a commercial flop, ResPower's render farm is busily working away right now. Over 700 machines. Over 11 million frames rendered. Anyone can use it. They accept credit cards. Good pricing, too.

    7054 frames in the rendering queue.
    7036 frames in the rendering queue.
    7001 frames in the rendering queue....

  52. Data, data, data by Xemu · · Score: 1

    Biggest issue with this service: Data. Fast data crunching needs data. Lots of data. How does one get it there? And the result back. It's expensive and slow.

    --
    Tell your friends about xenu.net
  53. Smaller Grid than they promosed? by HaydnH · · Score: 1

    From TFA:

    "Sun has spent the last 18 months trying to make this happen. First, it battled the basic logistics behind creating a massive publicly accessible cluster, then it battled the government. Regulators didn't care for the idea that rogue nations could log on to the Sun cluster and design nuclear weapons or run dirty bomb simulations in their spare time."

    "Sun has managed to assuage the government's fears by making the initial Sun grid a US thang only."

    "Sun does plan to open the grid up to international customers one day."

    "The grid that US customers find will be less ambitious than the one originally promised by Sun. Instead of myriad centers with 5,000 dual-core CPUs each, Sun will open one center with "less than" 5,000 dual-core CPUs."


    So they were planning for it to be international with 5000 CPUs... and now it's US only and it's less than 5000 CPUs? Wow now that's what I call a suprise!

    --
    Time is an illusion. Lunchtime doubly so. - Douglas Adams
  54. Linux yes! by Torg · · Score: 1

    The grid itslef is comprised ov V20Z servers. Depending on which datacenter you are talking about they are either dual 2.0Ghz or dual 2.2Ghz, each with 4G of RAM. They have no disks and are booted via PXE.

    The racks themselves have 32 systems, 4 racks make a pod (there was some discussion to call 2 rakcs a podlet). Depending on which datacenter you are talking about there are up to 3 pods.

    There are also some management servers that are really not worth mentioning. They serve as the infrastructure to boot the pods etc.

    One thing that is not put out on Sun's site is the pods will (and do) run Linux. You do not *have* to use Sol10 it is just that there are no Linux Infiniband drivers for the Infiniband (and yes thats about 2 tons of infinband cable connecting them, well ok in one data center).