Do you have GPU-brains? A poster-initiative.

This is a message to GPU-programmers only.

It is a simple question, and has many answers: what are GPU-brains? How is it possible your brain can code GPUs and only few friends and colleagues understand what you are doing? Is it thinking in parallel, focusing on one kernel and having the architecture in the back of the head. Is it simple loop-unrolling? Is it a web of thoughts? Is it just cool, as not many people can do it?

What to do?

I am asking you to draw or describe what GPU-programming is according to you. Around 9 stories/drawings will be used to make a poster (not based on above artist impression), of which 4 places already have been taken. If you are chosen to be on the poster, we’ll send you the poster and you’ll get eternal fame – this could increase the chance on a new job in GPU-development in case you are looking. We chose not to reward chosen texts/drawings with goodies, as

There is only two rules. 1) Don’t mention technologies like CUDA, OpenCL, OpenMP, AMP, etc. 2) You must have good experience with vector-processors and/or GPUs. Reason for the first rule is to prevent promoting a specific technology, and to make the poster less technical. The second rule is to make sure the poster is made by GPU-programmers. After seeing the poster, your boss must say to you: “Ah, I understand you’re cool and I should give you a raise”.

What will happen with your contribution?

The copyright of the poster and its content goes to the StreamHPC – this is to prevent others making variations which puts all text out of context. The poster can later be downloaded and freely be distributed – one is allowed to adapt the lower 10% to add one’s own logos. The goal is to promote GPU-programming, not to promote our company, a specific GPU-manufacturer or a specific technology. At StreamHPC we intend to use it at conferences and exhibitions.

The reason we want some control, is that your name is on the poster and we don’t want your contribution to be placed out of context. If you have better ideas, please let us know!

Send it!

You can send your text or drawing to poster[at]StreamHPC[dot]eu (private) or put it in the comments (public). Important! By sending your creation, you give StreamHPC all rights to be able to make the poster and derivative work, and to distribute the work under any creative commons license.
Also tell in the mail/comment how you want to be mentioned (twitter, full name, linkedin, etc). If you are looking for a job, also send your CV.

Related Posts

    We sponsor HiPEAC again this year

    ...  I find this conference very important. Compilers have the job to do magic. Last year I had nice feedback on my request to ...

    OpenCL – the battle, part III

    ...  While Tetris will live forever, the latest games also have something to add: realistic physics simulation. And that's what's done ...

    Disruptive Technologies

    ...  the evolved OpenCL 2.0. X86s get hardware-extensions to have ARM-support., to be able to run both "mobile widgets" and "power apps" ...

    OpenCL at SC14

    ...  Carine Morel, Fabrice Vidal 391-92 17:15 - 19:00 Poster GPU Acceleration of Small Dense Matrix Computation of the ...

3 thoughts on “Do you have GPU-brains? A poster-initiative.

  1. pip010

    “After seeing the poster, your boss must say to you: “Ah, I understand you’re cool and I should give you a raise”.” ahhh havent read an article so full of cliches recently 😀

    • StreamHPC

      Hehehehe. Well, this is a non-technical article to promote GPU-programming. One of the current obstacles is that managers don’t put understand what GPUs can do for the company.

      And what is more cliché than a manager who tells you what to do, what you suggested him last year?

  2. Brian Cole

    My graphical skills are limited, but I do have a story to tell that I think embodies my thought process as a parallel programmer.

    I was at a SuperComputing conference tutorial session one year waiting in a really long line to get a mid-day snack from the table full of goodies. Upon arriving at the table I realized half the table was full of desirables everyone wanted and the other half wasn’t as desirable. The desirable half of the table only had one set of tongs, while the undesirable half of the table had two sets of tongs. I spotted the clear tong load balancing problem and used one of the unused tongs to double the throughput of the snack table. The guy behind me exclaimed, “Parallelism!” There must of been dozens of HPC programmers in front of me, why didn’t they perform the same optimization?

Comments are closed.