[clean-list] Matrix timings
Shivkumar Chandrasekaran
shiv@ece.ucsb.edu
Mon, 29 Oct 2001 11:24:40 -0800
Recently Siegfried Gonzi wrote:
> Timings (on a Macintosh 100MHz, 603e processor, 64MB RAM, 16K L1, 256K
> L2):
>
> Beeing anti-verbose:
>
> Roughly speaking: takes for a 300x300 array and a vector set of array
> dimension 300x300
> about 20sec.
>
> The comparable Yorick execution time is about 15sec.
This timing is way too slow. On my PowerBook G4 (400Mhz) (running Mac OS
X), Scilab took 0.31 seconds to solve a system with 300 equations and
300 unknowns. Allowing for a processor that is 4 times slower we would
"expect" to solve it in 1.21 seconds on a 100MHz processor. Scilab (on
my machine) uses ATLAS BLAS. If we disallow for tuned BLAS
(conservatively) we can expect the 100MHz processor to be 16 times
slower, yielding an expected running time of 0.31 * 16 = 4.96 seconds.
In my own timings I found that Clean is typically 2.5 to 5 times (not
1.5 times) slower than equivalent code compiled with gcc. The best
performance (2.5 times slower than C) was obtained when I simulated 2D
arrays with 1D arrays.
--shiv--