IP: History recalled and some explanation re National Science Foundation to Fund Supercomputer

[David Farber <dave () farber net>]

> X-Sender: >X-Sender: catlett@localhost
> X-Mailer: QUALCOMM Windows Eudora Version 5.1
> Date: Fri, 10 Aug 2001 14:20:31 -0500
> To: farber () cis upenn edu
> From: Charlie Catlett <catlett () mcs anl gov>
> Subject: Re: IP: re National Science Foundation to Fund Supercomputer
>
> Hi Dave-
>
> In the past 24 hours, >In the past 24 hours, SETI@home has been smoking away at a hearty and 
> impressive 34 TFLOPS, about 3x what we are planning to put in place for 
> the TeraGrid.
>
> These are different types of capabilities, though, and while we can use 
> the same units (floating point operations per second) to measure them it's 
> a bit tricky to try to compare them, much less pit them against each 
> other.  Some applications will run wonderfully in a very widely 
> distributed environment, but others require more frequent, or higher 
> volume, exchange of data between the individual pieces.  We are really 
> trying to address these types of applications with TeraGrid, and we try to 
> look for places where we have common challenges and therefore can share 
> solutions with the innovative stuff happening in the >solutions with the innovative stuff happening in the SET@home 
> style grids.
>
> There was nothing sneaky intended with the "of its kind" wording- what we 
> are doing with the Terascale is very different from the good work of the 
> > SETI@Home folks, or distributed.net, Entropia, and others.
>
> But then as far as I can tell you wrote the first paper on grids, about 30 
> years ago.  So I guess I am preaching to the Pope here!
>
> Best Regards-
> Charlie Catlett (DTF network architect), Argonne National Lab



For archives see: http://www.interesting-people.org/