Author: Robert Hyatt
Date: 10:55:18 11/14/97
Go up one level in this thread
On November 14, 1997 at 08:25:18, Bas Hamstra wrote: >On November 11, 1997 at 09:42:55, Robert Hyatt wrote: > >[asm faster than C] > >>However, there is *much* more to the topic. *if* you are willing to >>"write" >>assembly language code, rather than just re-writing C code into assembly >>code, the gains can be much bigger. One example: Cray Research has >>what is >>considered to be the best FORTRAN compiler on the market, for obvious >>reasons. >>In Cray Blitz, we spent a couple of years writing assembly code and >>rather >>than a paltry 20-40%, we got 500% faster. Why? Because we didn't >>"convert" >>FORTRAN to assembly, we completely re-designed the program with the >>underlying >>architecture firmly in mind. > >Five years ago you could gain 300% or 400% on X86 C compilers too, by >doing >it in assembler. But compilers have improved significantly since then. >Based on my own (limited asm) experiences I find it hard to believe that >400% would be possible with todays compilers (Visual C/Watcom even GNU). > >Maybe it would be fun if you published the specs a for a very simple >and very short benchmark test here (I hope you can find the time). >Then we could all do our best jobs in asm and C and see how they >compare. > >And that test, not only simple to you but to anyone, please :) > > > >Best regards, >Bas Hamstra. That's part of the problem. "simple" tests get eaten by the compilers quite nicely.. The biggest gains are in complex tasks where you actually tailor the algorithm to the architecture, rather than to the programming language. But reasonable gains can be had just by exploiting knowledge that you have but the compiler doesn't, about the way the code executes...
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.