[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: CLX Speed: Pretty poor

Your message:

    We ran some tests to see how fast CLX is, and it isn't.    Compared to
    calling the X routines from C, CLX is **approximately 15 times slower**
    (1500% slower).  This is running the standard CLX under CMU CommonLisp
    on IBM RTs with X11R3.  My impression is the Lucid CLX is not faster (and
    may be even slower).  

    * Is CLX on some other machines faster?  (I challenge you to execute this
    	test yourself and see!)

    The Moves/Second is the number of rectangles that can be drawn (or
    drawn and erased) per second (so bigger numbers are better).
    The Bytes/Move is the number of cons
    cells allocated (why are there any at all?).  Only two of the tests have
    been run in C so far.

    					Moves/Second	Lisp/C Comparison
    	Moves/Second	Bytes/Move	    in C	(moves/sec)
    	------------	----------	--------------	-----------
    Test1	    129.5	    32		     1860.5	   x14.37
    Test2	    930.0	    32
    Test3	    114.9	    40
    Test4	    104.7	    64
    Test5	     89.1	    80		     1465.2	   x16.44
    Test6	     72.2	   144  

    Here follows the Lisp code which produced these results:

Does the code you sent correspond to test 1?

I changed the "display-finish-output" to "display-force-output" and ran it
on a sun4/260 using Franz Allegro CL 3.1beta, and May Day CLX.  Display
was "unix:0".

I got about 3500 "moves"/cpu second, and no consing.

	-- John Irwin, jdi@franz.com