Parallel – Super all

March 1st, 2013 by admin software


slidecast, Adateva founder and CEO Andreas OLOFSSON describes a new parallel project, the open source development environment for low-power parallel computing. In seeking to finance the project, the company’s start parallel to Kickstarter campaign in order to provide a development platform in just dollars. For more information, visit: www.kickstarter.com “inspired by a major hardware of the Raspberry Pi and Arduino, we see a critical need for a truly open, high-performance computing platform that close, lack of information in parallel programming,” said Andreas Olofsson, CEO of Adapteva. “High-performance parallel computing is currently only available in a small and select group of expert programmers to parallel project democratize access to parallel computing. Parallel programming problem is a major challenge of computer science, and we are all hands on deck.” Read more at adapteva.com


Part one of a two-part video well-known theoretical result in computer science: to stop the problem.
Video Rating: 4/5

Be Sociable, Share!

Leave a Reply

Security Code:

47 Responses to “Parallel – Super all”

  1. Comment by Tate A

    @TheMaster3P It would be crap unless the OS and applications are written to take advantage of all of the cores in parallel. If the programs are written in this way, the board can compute huge amounts of data quickly, otherwise, it will be slow.

  2. Comment by GeorgeChar95

    imagine playing gta V on this

  3. Comment by TheMaster3P

    But how powerful is this? Say, a 16-core 99$ board how powerful will be,compared to a low-end desktop pc?

  4. Comment by softilol

    Now i get it :D

  5. Comment by John Kang

    Aha! The problem nowadays is not that we have multiple cores but no one really know how to use them all. With Parallela, you can easily implement salable parallel computing in programming level.

    What does this mean for you? Supercomputers are obvious ones, also your normal everyday computing for under 1 Watt, our typical desktops use 300Watts under load!-and that’s not a gaming rig. If this thing really lifts off, we will see smarter smartphones, powerful game consoles, etc.

  6. Comment by softilol

    yea i know, the atoms are arranged in 2D plane… i am not illiterate lol

  7. Comment by dnbmania

    Of course it will keep on getting smaller, but the increase of cores is such a obvious way forward in power. Would you rather have 1 man trying to dig a mine or 1000? If programming can catch up it’ll scale so well. And heat really shouldn’t be an issue in the future where we’re heading towards more power for less power, hopefully that’ll keep on track so heat is negligible. And just to say, nothing is truly 2D ;)

  8. Comment by softilol

    part 2: with amazing properties. we can even stack 2d molecules one on the top of each other to create new properties. electrons move thru them near the speed of light. new element, silicine, is similar to graphene, but it has better compatibility with exsisting tech. there is lots of research to be done, and there is allready many, many people researching those and even more 2D molecules.

  9. Comment by softilol

    i dont think in the same way. multiple core-ing if i may express like that, became as a result of not being able anymore to make smaller processors, that are faster and also dont heat up so much.
    i think the future are processors on atomic scale. that is, atomic scale transistors. few years ago, we discovered graphine. if you dont know what it is, i strongly recommend you to look it up, very interesting and amazing molecule. its 2D, and now all sorts of new 2D molecules are discovered with

  10. Comment by dnbmania

    They only struggle to use multiple cores because that’s the way we’ve developed processors. As everyone WILL begin to change, it will become easier to access multi-core power even for low level programmers. programs like Visual Basic will optimise code for multi-core use. It’s the logical and definite solution for the advance of computers.

  11. Comment by Forget43

    “As soon as they get that running for 1 milli-second, other computers won’t matter.” Except that quantum computers require absurd cooling, to near absolute zero and offer very little performance as of now. (I know they are stronger in other types of calculations, but they’re even suffering in those calculations if you search for info about it.)

  12. Comment by Dev Jag

    no contact on your website. where to order?

  13. Comment by softilol

    “LOL”, “OMG”, “ROFL”, “WTF” + *insert random insult here* + no explanation = the legit argument on the internet, that makes you look smart.

  14. Comment by binary132

    LOL. dumbass. that’s obviously the point…

  15. Comment by generischernutzer

    Well, yeah. 640kb ought to be enough for anybody. ;-)
    (Don’t mind that quote probably being taken out of context.)

  16. Comment by Kevin Huynh

    i am learning programming and my teacher assigned a project and i became facinated by the parallella and would like the power point shown in the video if possible.

  17. Comment by Vasilian Sotirov

    What’s the song at the beginning of the clip?

  18. Comment by lifeisgood070

    Mass production of the computer chips we have now is super cheap… you obviously don’t spend time in the electronics industry. If you did, you have to sign numerous agreements. But most chips are dirt cheap to manufacture and the raw materials are also dirt cheap. The cost is high, because 1) companies get away with it & 2) they include a lot of other stuff in the budget like extensive engineering, failed projects, research, other funding etc. Proof is in the profitability of $100 warrentys

  19. Comment by softilol

    And yes, mass production of graphine is super easy and actually CHEAP! Also other two dimensonal molecules are showing their amazing OUTSTANDING different specifications and abilities.

  20. Comment by softilol

    …easy to manage spraying oxygen atoms on it to make it the shape you want
    ( graphane ) and it even repairs itself in the way, that he sucks the carbon atoms that are close and fill it in a gap! And it can be even used as a filter because its pores can only let water thru. Perfect for cleaning water !. But for computing, there will be processors with about 150Ghz.

  21. Comment by softilol

    Noob. please stop replying. Quantum MAY worked if the information(s) that electon carries wouldnt CHANGE every-fucking-time the info is READ. But i wont argue, if they get it running, its beast. You are actually right.
    People are not “STRUGGLING” to make it small. Its super easy to make graphine. Actually you are making ( unintendintly and in small quantaties ) it even when you write with a pencil :) . Its super powerful, electrons move thru it like it was nothing there at all, transparent, …

  22. Comment by lifeisgood070

    I live on this planet mr. nay-sayer. Go be grumpy by yourself. There’s more than just this molecule bs transistor stuff. Have you taken a look at IBM’s quantum computer? As soon as they get that running for 1 milli-second, other computers won’t matter. Besides… you are proving your own point. If people are struggling to make things so small, why not change the focus to parallel computing & make it more common place. Psh I remember when i lived under a bridge…. troll.

  23. Comment by softilol

    On witch planet do you live ? First study 2 and 4 cores, and how programmers struggle to use them, then say this kind of stuff. What we will get is molecule-based transistors ( the physicial limit of how small can it be ) by two dimensional molecules like graphine ( graphane offers huge possibilities for computing ).

  24. Comment by lifeisgood070

    Those are the kind of questions that stop progress. I mean my computer runs internet explore, why would I want more? Maybe someday we will all do super-duper-scientific calculations and animations. Join Folding@Home…. there are tons of cool things to do with 64 cores.
    I just wanna know how they get that 45ghz number….

  25. Comment by JoshDangerFabean

    I wasn’t questing how processors work. I was simply saying why wouldn’t you want more processors. Nothing is hurt from having more processors. Especially if they’re giving you a processor for so cheap.

  26. Comment by hbkrulestheworld

    if you already know the problem skip to part 2 :D

  27. Comment by foldthirteen

    Awesome delivery of the concept, truly awesome.

  28. Comment by ogamer666

    1 like for every 100 views :D

  29. Comment by elos12

    That is one powerful computer monitor

  30. Comment by rajohns08

    what’s that blue square eve hands bob?

  31. Comment by Bujiraso

    Robots won’t take over.
    Hate to burst everyone’s bubble but only humans can figure out the answers to certain problems which MUST be solved in order to program a computer.
    And we’re not just talking it’s hard, we’re talking it’s an established mathematical fact that has been pondered and worked on for about 76 years now.
    Type “The Halting Problem” into youtube and watch the first link

  32. Comment by tekal85

    Bobs proposed solutions are the funniest part of the whole story.

  33. Comment by Quinn Culver

    Though there is indeed an algorithm that solves the halting problem for the finite collection on the disk, I don’t think anyone could write it, since they’d have to hard-code the solution. They could, actually, but only buy guessing correctly. However, saying that “it would take a long time to write” seems to indicate that it’s figure out-able, which it’s not (i.e. the solution is not uniform in the finite set).

  34. Comment by Quinn Culver

    @stoobers1 Notice that what I’m suggesting would be impossible if there were infinitely many programs on the disk, since an algorithm cannot hold infinitely much hard-coded information.

  35. Comment by Quinn Culver

    No, I’m saying there’s a program that “guesses” correctly, since only finitely many guesses are needed. E.g. if there is only one program, P1, on the disk, then either P1 halts or it doesn’t, so two guessing algorithms are sufficient; one that outputs “yes” (no matter what) and one that outputs “no”. One of these algorithms is correct. Similarly if there are n programs on the disk, then 2^n guessing algorithms (on for each combination of “yes” and “no”) are sufficient.

  36. Comment by stoobers1

    I think I understand. So if the program you are testing on your finite inputs halts, then hooray. But if the program is taking forever, you could look at the loop conditions and the current input state and determine if the loop will quit or not. Is that right?

  37. Comment by Quinn Culver

    The disk has finite memory, hence the collection of (indices of) programs on the disk is finite. Every finite set is computable. QED

  38. Comment by stoobers1

    Prove it.

  39. Comment by ClericAgent

    Eve is such a bitch.

  40. Comment by boumbh

    Actually, the problem as stated in this video is solvable… If you take a finite set of programs, you can always create a program that says true or false whether the input program is flawed or not. The number of programs which size is < 1To is finite. Therefore, you could even write a program that checks all the programs < 1 To. It would take a long time to write though XD .

  41. Comment by manrayer88

    Why does bob have a Mohawk
    Punk rockers aren’t educated enough to ponder this sort of thing
    and why is eve shirtless that dirt whore

  42. Comment by Quinn Culver

    There is indeed an algorithm which, given input “n”, outputs “yes” if the nth program on the disk halts and “no” otherwise. Bob just can’t find it.

  43. Comment by kjlg74

    That’s the kind of reaction I like to hear :D

  44. Comment by fractal420

    i must know!

  45. Comment by kjlg74

    It’s a gripping cliffhanger, isn’t it? ;D

  46. Comment by kjlg74

    It’s anyone’s guess…

  47. Comment by 1GOD1JESUS

    OMG what is going to happen next?

Get Adobe Flash player