andrel wrote:
> Anyway, the best we can hope for is a continuation of POV along the 
> lines that we are currently following.

No, the best we can hope for is the introduction of heterogenous arrays 
of processors whith automatic load balancing and sharing.

Such processors would automatically vectorize the code run on them when 
appropriate; in other words, programs wouldn't need to be rewritten to 
use new libraries, they would just run faster.

It's a similar concept to how out of order execution works; the 
processor examines the code, rearranges it and sends it to the various 
processing units as appropriate.

...Chambers

Post a reply to this message