Re: Are 64 Int or FP registers useful?

>neideck@nestvx.enet.dec.com (Burkhard Neidecker-Lutz) writes:
>>Most compilers couldn't deal with that many registers. A good paper about
>>a compiler system that could is:
>
> [References to WRL technical reports deleted.]


and hrubin@pop.stat.purdue.edu (Herman Rubin) writes:
>With all these claims about how tremendous a job compilers are doing, this
>is certainly a counterexample.


The first posting makes a bad start and the followup leaps on the first
statement and ignores the counterexample given by the refs. I'd rewrite
the first statement as


Some compilers would make faster code if they would
use more registers.


Probably we can extend "some" to "most", since most compilers are entirely
without optimization (counting all the academic experiments).


Regarding 64 registers...


I believe that 64 FP registers is probably past the point of diminishing
returns for most application programs. On the other hand, there are
important examples that could profitably use more than 32 registers.
Further, I believe that as memory latency increases (in terms of FP
speed), the number of useful registers will increase.


For an important example, I'd cite dense matrix algebra packages. Of
course, if your primary applications are integer based (say, you edit and
compile all day), extra FP registers aren't going to help much.


Preston Briggs
--

Return to the comp.compilers page.
Search the comp.compilers archives again.