[Beowulf] Berkeley View: A future in Parallel Programming?

[Peter St. John]

I'm really enjoying the PDF there,
http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-183.pdf,
particularly, the part about optimizing networks for applications. But the
nice general interest takeaway is that list of Old Conventional Wisdom
paired with New, that's fun. Thanks.
Peter
Old CW: Power is free, transistors are expensive.
New CW: (from the pdf) The transistors are free, but the power is expensive
because you can't afford to power all the transistors on the chip.
Synthesis: Power and Transistors are free, but Density is expensive.


On 3/15/07, Thomas H Dr Pierce <TPierce at rohmhaas.com> wrote:
>
>
> Dear Beowulf ML,
>
> Here is an interesting discussion on the methods and metrics that could
> apply to multicore chips and clusters.
>
> I have not seen this discussed on the list, so it may be new to some.
>
> Here is the link to the overview wiki
> http://view.eecs.berkeley.edu/wiki/Main_Page
>
> And motivation for people to go to the link: quote from the wiki
>
> "We believe that much can be learned by examining the success of
> parallelism at the extremes of the computing spectrum, namely embedded
> computing and high performance computing. This led us to frame the parallel
> landscape with seven question under the following assumptions:
>
>    - The target should be 1000s of cores per chip, as this hardware is
>    the most efficient in MIPS per watt, MIPS per area of silicon, and MIPS per
>    development dollar.
>    - Instead of traditional benchmarks, use 7+ "dwarfs<http://view.eecs.berkeley.edu/wiki/Dwarfs>"
>    to design and evaluate parallel programming models and architectures. (A
>    dwarf is an algorithmic method that captures a pattern of computation and
>    communication.)
>    - "Autotuners" should play a larger role than conventional compilers
>    in translating parallel programs.
>    - To maximize programmer productivity, programming models should be
>    independent of the number of processors.
>    - To maximize application efficiency, programming models should
>    support a wide range of data types and successful models of parallelism:
>    data-level parallelism, independent task parallelism, and instruction-level
>    parallelism. "
>
>
>
> And the detailed white paper that  started his (Dec 2006)
> http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-183.html
>
> The "Seven Questions" discuss approaches to what seem to be standard
> discussions on cluster, parallel programming and best practices.  Lots of
> fun for everyone!
>
> Applications
>
> 1. What are the applications?
> 2. What are common kernels of the applications?
>
> Architecture and Hardware
>
> 3. What are the HW building blocks?
> 4. How to connect them?
>
> Programming Model and Systems Software
>
> 5. How to describe applications and kernels?
> 6. How to program the hardware?
>
> Evaluation
>
> 7. How to measure success?
>
> ------
> Sincerely,
>
>   Tom Pierce
>
>
>
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.beowulf.org/pipermail/beowulf/attachments/20070316/4c558cbf/attachment.html>