Re: C++ programmer and assembly
allnor@tele.ntnu.no (Rune Allnor) wrote (abridged):
Until recently, I have structured my C++ programs based on the
point of view that what I learned of assembler programming many
years ago is still valid.
What you learned many years ago may be invalid even if you were working in
assembler directly. Hardware improves, and the relative costs of
operations change. For example, often now some operations are effectively
free because they can be done in parallel with other operations.
On a couple of occations I have recieved advice here which can best
be summarized as "forget everything you knew back then, program
straight-forward code and let the STL and the compiler take care
of performance issues."
That's a good way to write fast code, provided you use the time saved by
not doing optimisation in the majority of cases to optimise really well
the few cases that actually matter. And provided you pay attention to the
higher level algorithms.
Assembly level optimisations are generally low-level. They can sometimes
improve performance by a significant linear factor, but they rarely
improve from O(N*N) to O(N).
-- Dave Harris, Nottingham, UK.
--
[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]