[Jprogramming] Aiming at a multi-core future

[Don Guinn]

They will probably use them all. But isn't the one available on PCs for less
than $1000 the shared main memory? A real market there. And it's true
that there are many cases which are very difficult to coordinate with
multiple processors. But there are some which are easy. They are a good
place to start. We should not let the cases that are hard prevent the
investigation of the cases which are easy.

On Dec 20, 2007 11:27 AM, Steven H. Rogers <steve at shrogers.com> wrote:

> Don Guinn wrote:
> > As to moving data around between various memories - I am not really
> clear on
> > the architecture of multi-core systems, but don't they consist of
> multiple
> > processors, each processor having its own cache but sharing a common
> main
> > memory? If so, then there is not a problem of having to move blocks of
> data
> > among different main memories.
> There is  a problem moving blocks of data between different chunks of
> memory.  It may be possible to hide the problem from application
> programmers some, or even most, of the time, but it's there and it has a
> cost.  Partitioning a problem among available memory and processors
> efficiently is often non-trivial.
>
> There is not one multi-core architecture, but many.  Different degrees
> of parallelism and different workloads can drive different architectural
> decisions.  Processors may be symmetric or asymmetric, with local and/or
> shared memory.  Processors may communicate via shared memory, buses, or
> crossbar switches.  It isn't clear which architecture(s) general purpose
> computers will use.
>
> # Steve
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
>