[S-News] Re: S-PLUS on multi-processor machines

Jens Oehlschlaegel (oehl@Psyres-Stuttgart.DE)
Thu, 5 Feb 1998 12:50:17 +0100 (MET)


I am wondering why the discussion on the multi-processor issue is mainly
focusing on how to run several S+ processes in parallel.

Isn't S+ a vector oriented language? Isn't a vector operation just the
ideal condition for parallelization?

S+ has memory and performance problems in those tasks, which *cannot* be
written in vector operations, i.e. where processing of one vector element
dependes on processing of the previous vector elements: clearly a
situation where parallel processing is not feasible.

So every S+ programmer tries to formulate their tasks in vector
operations, where elements can be processed independently of other
elements, and thus parallel processing is feasible: ideal conditions for
parallel processing.

It would be a pitty, if S+ on a multi-processor machine would not make use
of these ideal conditions, and all our efforts in writing vector code.

Am I wrong?

Best regards

--
Jens Oehlschlaegel-Akiyoshi
Psychologist/Statistician
Project TR-EAT + COST Action B6
                                                 F.rankfurt
oehl@psyres-stuttgart.de                         A.ttention
+49 711 6781-408 (phone)                         I.nventory
+49 711 6876902  (fax)                           R .-----.
                                                  / ----- \
Center for Psychotherapy Research                | | 0 0 | |
Christian-Belser-Strasse 79a                     | |  ?  | |
D-70597 Stuttgart Germany                         \ ----- /
-------------------------------------------------- '-----' -
(general disclaimer)                             it's better

----------------------------------------------------------------------- This message was distributed by s-news@wubios.wustl.edu. To unsubscribe send e-mail to s-news-request@wubios.wustl.edu with the BODY of the message: unsubscribe s-news