Re: more fun with proposals

Eric Salo (salo@mrjones.engr.sgi.com)
Thu, 11 Jul 1996 15:15:01 -0700

> For example, splitting is not feasible if:
>
> . you have a large region ( hundreds of megabytes) that you access in small
> sections (kilobytes or less) possibly with variable size and shape
>
> . you cannot decompose the region into multiple windows because
> you don't know in advance how (might depend on dynamic load balancing)
>
> At the same time, the application is able to guarantee that there
> is no concurrent overlapping acceses, for example due to the physical nature
> of the problem -- "if I work on molecule A, then there is no any other
> process that could possibly update the array element potential(5, 9001,
> 6107)", and is not willing to pay the cost of locking.

Ah, I think I understand now. You've got some huge amount of data/work that is
being distributed between different processes, and each chunk of work has an
associated data structure. You don't know in advance which process will be
working on a particular chunk, but do know that only one process at a time will
need to access the data structure for each chunk (at least, until the end of
the current iteration). So while you *could* create a seperate window for each
data structure, it would be highly impractical in the case where you had tens
of thousands of them.

Thanks for clearing that up.

-- 
Eric Salo         Silicon Graphics Inc.             "Do you know what the
(415)933-2998     2011 N. Shoreline Blvd, 8U-808     last Xon said, just
salo@sgi.com      Mountain View, CA   94043-1389     before he died?"