Sujet : Re: "Mini" tags to reduce the number of op codes
De : anton (at) *nospam* mips.complang.tuwien.ac.at (Anton Ertl)
Groupes : comp.archDate : 08. Apr 2024, 08:16:08
Autres entêtes
Organisation : Institut fuer Computersprachen, Technische Universitaet Wien
Message-ID : <2024Apr8.091608@mips.complang.tuwien.ac.at>
References : 1 2 3 4 5 6 7
User-Agent : xrn 10.11
Thomas Koenig <
tkoenig@netcologne.de> writes:
But beating RISC-V is easy, try getting you instruction count down
to VAX counts without losing the ability to pipeline and parallel
instruction execution.
>
At handwaving accuracy::
VAX has 1.0 instructions
My 66000 has 1.1 instructions
RISC-V has 1.5 instructions
>
To reach VAX instruction density
Note that in recent times Mitch Alsup ist writing not about code
density (static code size or dynamically executed bytes), but about
instrruction counts. It's unclear why instruction count would be a
primary metric, except that he thinks that he can score points for My
66000 with it. As VAX demonstrates, you can produce an instruction
set with low instruction counts that is bad at the metrics that really
count: cycles for executing the program (for a given CPU chip area in
a given manufacturing process), and, for very small systems, static
code size.
- anton
-- 'Anyone trying for "industrial quality" ISA should avoid undefined behavior.' Mitch Alsup, <c17fcd89-f024-40e7-a594-88a85ac10d20o@googlegroups.com>