Integer Multiply/Divide on Sparc
Teemu Torma
tot at frend.fi
Tue Jan 2 03:21:24 AEST 1990
In article <1313 at kuling.UUCP> irf at kuling.UUCP (Bo Thide') writes:
================================================================================
register auto auto int function auto
int short long multiply call+ret double
HP9000/370 (fpa -O) 0.22 0.26 0.22 1.35 3.96 0.62
HP9000/370 (-O) 0.21 0.26 0.22 1.35 3.08 1.21
HP9000/370 (fpa no -O) 0.26 0.40 0.36 1.44 4.42 1.56
HP9000/370 (no -O) 0.26 0.40 0.37 1.45 3.38 2.72
HP9000/835 (-O) 0.27 0.29 0.27 5.49 0.31 0.27
HP9000/835 (no -O) 0.29 0.53 0.45 5.62 0.31 0.59
Sun SS1 (-O) 0.29 0.33 0.30 19.5 0.49 0.59
Sun SS1 (no -O) 0.38 0.40 0.35 19.7 0.51 0.72
================================================================================
There is no question that Sun SparcStation 1 is *extremely* slow on integer
multiply even for a RISC architecture -- scaling the HP-PA results to the
same clock speed as the SPARC (20 MHz) we see that HP-PA is about 470% faster
than SPARC!! Our results also show that integer arithmetics on CISC (MC68030)
is much faster than on RISC (HP-PA, SPARC).
Strange. I got much better int multiply results from Sun SS1. Gcc
version was 1.36.
register auto auto int function auto
int short long multiply call+ret double
Sun 4/60 (cc, no -O) 0.38 0.40 0.36 3.52 0.28 0.72
Sun 4/60 (cc, -O) 0.32 0.35 0.32 3.62 0.28 0.62
Sun 4/60 (cc, -O2) 0.30 0.33 0.30 3.32 0.26 0.61
Sun 4/60 (cc, -O3) 0.30 0.33 0.30 3.45 0.28 0.63
Sun 4/60 (gcc, no -O) 0.21 0.38 0.38 3.50 0.27 0.77
Sun 4/60 (gcc, -O) 0.18 0.21 0.18 3.57 0.27 0.42
Sun 4/60 (gcc, <2>) 0.17 0.22 0.19 3.67 0.28 0.42
Sun 4/60 (gcc, <3>) 0.17 0.21 0.18 3.52 0.27 0.40
gcc <2>: -fstrength-reduce -fcombine-regs -fomit-frame-pointer
gcc <3>: as <2> including -mno-epologue
--
Teemu Torma
Front End Oy, Helsinki, Finland
Internet: tot at nyse.frend.fi
More information about the Comp.lang.c
mailing list