Re: [oc] Processor Instruction reply for Andreas

[Johan Klockars <rand@cd.chalmers.se>]

> > - Depending on implementation, may take different number of
> >  cycles to execute, when a branch is taken, vs. not taken.
> 
> Look-ahead as implemented for years now automatically scans for
> jumps to make sure it has the 'alternative' piece of code ready
> if the jump is actually committed thus jumps don't 'surprise+delay'

Can you name a single processor that does this?
I have never heard of any.

> the pipelining. I think they even use 2+ integer pipelines in most
> processors these days maybe even for this reason?

Definitely not.

What modern processors do is to predict which way a branch is most
likely to go and then fetch instructions from there. If the prediction
turns out to be wrong, there is a often _huge_ performance hit.
A correctly predicted branch may take one, or even zero, cycles, while
a mispredicted one will need to clear the pipelines and thus can cost
tens of cycles.

> > For 2:
> > - Two instructions, need larger program memory
> 
> Given that not every instruction is suitable for conversion yes, tests
> show that on average a 60-70% increase in size. Instructions like
...
> faster one. I usually find it easier to add another DIMM than over clock
> my processor by 80%.

The external RAM is of course not a problem, but the caches may well be.

> > + faster execution
> 
> Mine? Of course, less clock cycles to do the same job means faster execution
> of the same task.

If you can manage to keep the cycle time up, and if you can still
make as good use of multiple pipelines.b, and if you do not use up
too much memory bandwidth...

> > To summaries, a RISC architecture typically tries to keep it's
> instructions
> > very simple and easy to execute. The goal for a RISC is to execute a
> > small number of instructions at a very high speed. This results in overall
...
> If anything my design depending on how you view it either reduces the
> instruction
> set or keeps the same amount of instructions. I'll have more details and a

The important thing is 'simple', not 'few' instructions.

-- 
  Chalmers University   | Why are these |  e-mail:   rand@cd.chalmers.se
     of Technology      |  .signatures  |            johank@omicron.se
                        | so hard to do |  WWW:      rand.thn.htu.se
   Gothenburg, Sweden   |     well?     |            (fVDI, MGIFv5, QLem)
--
To unsubscribe from cores mailing list please visit http://www.opencores.org/mailinglists.shtml