Re: Applecrate II on Slashdot

["Michael J. Mahon" <mjmahon@aol.com>]

datajerk wrote:
> On May 3, 4:22 pm, nyder <nyder...@gmail.com> wrote:
>
> > http://hardware.slashdot.org/story/11/05/03/1827247/AppleCrate-II-App...
> >
> > Your famous Michael J. Mahon, grats on your site possibly getting
> > slashdotted.  =)
>
>
> Awesome work Michael.  I'd love to see some benchmarks with one, two,
> four, ... nodes.

Thank you!

I reported some performance work (on the AppleCrate I) at:

http://home.comcast.net/~mjmahon/ParallelSim.html

As you would expect, performance scales pretty nearly linearly
as long as network traffic does not overload NadaNet.  This is
typically the case for "embarassingly parallel" problems which
can be split into computational parts that can execute essentially
without interaction (like Mandelbrot, or a Monte Carlo simulation,
for example).

And there is a more latency/bandwidth-related performance study
reported in the File Server article:

http://home.comcast.net/~mjmahon/FileServer.html

> Clearly the next step is AppleCrate III made from Apple ///s.  :-)
> But please don't do that since the /// is rare and I still do not have
> one.  :-(

The number has nothing to do with the II in Apple II--it's just
"Mark II", since the original 8-board AppleCrate preceded it (and is
now retroactively named AppleCrate I ;-).

-michael

NadaNet 3.1 for Apple II parallel computing!
Home page:  http://home.comcast.net/~mjmahon/

"The wastebasket is our most important design
tool--and it's seriously underused."