Re: Why VAX Was the Ultimate CISC and Not RISC

On Mon, 03 Mar 2025 07:39:03 GMT
anton@mips.complang.tuwien.ac.at (Anton Ertl) wrote:

 
And looking at my latest code size measurements
<2024Jan4.101941@mips.complang.tuwien.ac.at>, both armhf (ARM T32) and
riscv64 (RV64GC) result in shorter code than IA-32 and AMD64:
 
   bash     grep      gzip
  595204   107636    46744 armhf
  599832   101102    46898 riscv64
  796501   144926    57729 amd64
  853892   152068    61124 i386
 

I never measured size of gnu utilities, but my measurements of few
of my own embedded projects and of some microbenchmarks always gave
very different ratios.
That is, in my measurements T32 was also a champion among extant 32b/64b
architectures (extinct nanoMIPS was better), but i386 was MUCH closer
than in your figures above. Up to twice closer, actually.
It seems, newer gcc is much worse than older versions at generation of
compact i386 code.
Also in my measurement T32 was significantly denser than RV64GC,
although in case of RV I only did microbenchmarks.

One of my early measurements that I have bookmarked.
https://www.realworldtech.com/forum/?threadid=86001&curpostid=86094