🚀 go-pugleaf

In article <PCG.92Dec13164348@aberdb.aber.ac.uk>, pcg@aber.ac.uk (Piercarlo Grandi) writes:
|> rschnapp> Don't forget about out-of-order (i.e., dataflow) execution
|> rschnapp> techniques.  You can gain plenty of additional fine-grain
|> rschnapp> parallelism.
|> 
|> Ah, there are indeed many interesting tricks to support high levels of
|> micro parallelism *in the implementation*. To me the really interesting
|> problem is not how to do it, though, it is whether it is worth doing it,
|> and for which classes of codes.
|> 
|> I have yet to find evidence that *general purpose* codes have an
|> intrinsic degree of micro parallelizable operation (let's call it
|> superscalarity) that makes it worthwhile to have a degree of parallel
|> *instruction issue* within a single stream not greater than 2-4.

...which is why I pointed out a paper you might wish to refer to.
There have been several studies indicating that there is plenty of
capacity for fine-grain parallelism in common programs.

Most of the SPEC '89 programs, for instance, benefit substantially from
out-of-order execution.  Supposedly, these ARE representative of
"typical" programs.  One of the nice things about out-of-order
execution is that you get the benefits of improved instruction
scheduling without having to recompile.
-- 

...Russ Schnapp
BIX: rschnapp           Email: netcom!metaflow!rschnapp or rschnapp@metaflow.com
Metaflow Technologies   Voice: 619/452-6608x230;  FAX: 619/452-0401
La Jolla, California    Unless otw specified, I`m speaking only for myself!