Re: is C necessarily faster than C++

tbrannon (tbrannon@mars.mars.eecs.lehigh.edu) wrote:
: This guy in my lab keeps refusing to use C++ in our program intended
: to simulate somatosensory neural circuits because he says it is slower
: than C. My (uninformed) response was that most of what you see as
: overhead (ie, message routing, value accessing, type checking) is
: optimized away at compile-time.


: Any pointers to more empirical studies?


Yes. My Masters thesis, entitled "A Comparative Study of the Locality
Charateristics of an Object-Oriented Language", compares the cache
locality of two programs that perform the same function. One is
written in C; the other in C++. Both were compiled with the same
compiler. Hardware aquired memory reference traces of the two programs
were compare on the basis of cache misses. The results indicates that
C++ programs can have 2-3 times more cache misses. On today's machines
that translates into a significant performance difference. On tomarrow's
machines, the effect will be even more pronounced.


It should be noted that I was not trying to compare C and C++. The
objective was to compare object-oriented'ness with a traditional style.
The C++ program was excessively object-oriented, while the C program was
well structured. Thus the results are probably (and were intended to be)
a rough bound on the penalty of adopting an object-oriented style.
Naturally, YMMV.


Finally, the thesis is not on-line (yet). I am having troubles getting
the postscript file to print properly (it was generated on a Macintosh
and doesn't seem to print properly from UNIX). I would imagine you can
obtain a hard copy by contacting the Computer Science Department,
Brigham Young University, TMCB, Provo, UT 84601.


--


-Mark- mkgardne@uiuc.edu (mkgardne@cs.uiuc.edu)
--

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