🚀 go-pugleaf

Julien Pierre wrote:
> 
> In article <4jhbus$dga@nuke.csu.net>,
> ghansen@isc.sjsu.edu (Gregory L Hansen) wrote:
> >And the result is:
> >
> >Hansen Floating-Point
> >Machine   OS      Processor   Raw  Normalized  Relative  Compiler
> >--------  ------  ----------  ---  ----------  --------  --------
> >Power100  Sys7.5  PPC601/100  82s     82s          1     1,2
> >==================================================================
> >1: CodeWarrior 8
> >2: optimized for speed
> >
> 
> Self-builtPC    OS/2 Warp       Pentium/100     54s
> 
> Using Watcom 10.0 compiler and all optimizations
> 

Note that I spent any time looking at the code to optimize it... but
here are (unofficial, I am doing nothing official here...  my numbers COULD be wrong) R4.4k at 150MHz,
{1,2,4,8,16}xsR8k@90MHz, {1,2,4,8}xR10k@196MHz.

Indy  Irix 5.3	R4400@150 MHz   (1 Meg cache)  Raw 44.6  compiler: cc -O3 -mips2 test.c -o test.x  -lm

PowerChallenge 6.1 1xR8k@90 MHz (4 Meg cache)  Raw 21.27 compiler: cc -O3 -mips4 -64 -SWP:=ON -sopt
-Wk,-ur=4,-chs=2048 test.c -o test.x -lm

PowerChallenge 6.1 1xR8k@90 MHz (4 Meg cache)  Raw 19.49 (converted code to fortran to obtain real
optimization)

PowerChallenge 6.1 2xR8k@90 MHz (4 Meg cache)  Raw 10.34 (converted code to fortran to obtain real
optimization)

PowerChallenge 6.1 4xR8k@90 MHz (4 Meg cache)  Raw 5.53 (converted code to fortran to obtain real
optimization)

PowerChallenge 6.1 8xR8k@90 MHz (4 Meg cache)  Raw 2.96 (converted code to fortran to obtain real
optimization)

PowerChallenge 6.1 16xR8k@90 MHz (4 Meg cache)  Raw 1.70 (converted code to fortran to obtain real
optimization)

(fortran compilation line: f77 -extend_source -O3 -mips4 -64 -SWP:=ON -pfa keep
-Wk,-ur=8,-so=3,-r=3,-o=5 -mp test.f -o test.x )

(fortran source:

	program test
	real*8 sum
	integer i
	
	sum=0.0d0
	do i=1,10000000
	   sum=sum+dsqrt(dsin(dble(i)/4.2d0)*dsin(dble(i)/4.2d0) + dlog(dble(i)))
	enddo
	print *,'total: ',sum
	end 
)

PowerChallenge 6.2 1xR10k@196 MHz (1 Meg cache)  Raw 6.70 (converted code to fortran to obtain real
optimization, using R8k binary)

PowerChallenge 6.2 1xR10k@196 MHz (1 Meg cache)  Raw 7.19 (converted code to fortran to obtain real
optimization, using R10k binary)

PowerChallenge 6.2 2xR10k@196 MHz (1 Meg cache)  Raw 3.88 (converted code to fortran to obtain real
optimization, using R10k binary)

PowerChallenge 6.2 4xR10k@196 MHz (1 Meg cache)  Raw 2.24 (converted code to fortran to obtain real
optimization, using R10k binary)

PowerChallenge 6.2 8xR10k@196 MHz (1 Meg cache)  Raw 1.55 (converted code to fortran to obtain real
optimization, using R10k binary)

I repeat, all of these are UNOFFICIAL numbers, there is nothing OFFICIAL about them.
I compiled them on my own, ran them on hardware available to me.  They are provided as-is.

Joe


-- 
+-----------------------+----------------------------+     /
|     Joe Landman       | +1 810 615 2169  voice     |    /
|   Systems Engineer    | +1 810 478 3181  fax       |   /
| Silicon Graphics Inc. |        x 5-7670  corporate |  /
|   24155 Drake Road,   | landman@detroit.sgi.com    | /
|  Farmington MI 48335  |   Mail stop DMR 225        |/
+-----------------------+----------------------------+