Tony wrote:
Providing that there's a way to get a bunch of cheapo CPUs to do the
job.
True, but there's a complexity hit (which translates into a reliability hit
and a development lag) from this approach.
Well, i do not think that using off-the-shelf cheapo CPUs debugged by
somebody else and tested in millions of applications, instead of
single-purpose ASICs qualifies as a reliability hit or a development lag
--vadim