Parallelizing Translation

Douglas Eadline deadline at
Wed Jul 5 19:32:46 PDT 2000

On Wed, 5 Jul 2000, Walter B. Ligon III wrote:

> I think the discussion at hand is a fine example of the fact that the every
> day parallelization techniques that are used for SO many important problems
> are NOT the only techniques out there.  The Beowulf movement has had the
> problematic effect that many self-proclaimed experts are simply unaware of
> much of the research in parallel computation that has gone on over the last
> 40 years - thus leading to miunderstanding and sweeping statements that are
> not really productive in the long run.
> Beowulf is not the untimate answer - and it will not replace all other forms
> of parallel computation.  There ARE problems that fit better onto other
> architectures.  Those architectures may not make economic sense right now,
> thus we may not see much of them for a while, but that does not invalidate
> their inherent value.

This is the most sane thing I have heard in a while.


Paralogic, Inc.           |     PEAK     |      Voice:+610.814.2800
130 Webster Street        |   PARALLEL   |        Fax:+610.814.5844
Bethlehem, PA 18015 USA   |  PERFORMANCE |

More information about the Beowulf mailing list