On Sat, Jun 22, 2002 at 04:57:49PM +0200, Dieter N?tzel wrote:
> On Saturday 22 June 2002 16:15, Jens Owen wrote:
> > Zilvinas,
> >
> > Thanks for posting this data.  It's nice to see confirmation of how
> > things were intended to work.
> 
> I'll second that.
> May I ask if Zilvinas could repeat with 60Hz?
> I know that this is a "bad number" but the RAMDAC clock has some influence.
> Most "official" benchmarks (Win) are made @60 Hz...

Ok if there is a interest :) in that data I see no problem ;) I will do
that in a nearest future ;) I am wondering myself what (if any)
difference that would make ;)

> 
> Second:
> My dual Athlon MP 1900+ (MSI K7D Master-L, AMD 760MPX) with 512MB DDR-SDRAM (1
> GB is comming soon) is up and running.
512MB RAM
ABit KT7A-RAID (KT133A chipset)
Single CPU (sighs. ...) Athlon XP 1800+ SDRAM (PC133).
Radeon 7200 (QD model which is VIVO DDR 64MB ... another name for that.)

> 
> I can get my hands on a Radeon 8500 and maybe a 7500, tomorrow or Monday.
> I'll do some viewperf-6.1.2 numbers on my V5 5500 @24 bit tonight.

I assume that these runs are done on Radeon (or it is V5 5500 ?)

> 
> Mesa/demos> ./gloss
> 596 frames in 5.006 seconds = 119.057 FPS <-- cylinder
> 764 frames in 5.003 seconds = 152.708 FPS
> 760 frames in 5.006 seconds = 151.818 FPS
> 758 frames in 5.001 seconds = 151.57 FPS
> 
> 483 frames in 5.004 seconds = 96.5228 FPS
> 420 frames in 5.003 seconds = 83.9496 FPS
> 420 frames in 5 seconds = 84 FPS
> 421 frames in 5.008 seconds = 84.0655 FPS <-- teapot
That's very interesting ...

$ ./gloss 
radeonUpdatePageFlipping allow 1 current 0
radeonUpdatePageFlipping allow 1 current 0
1485 frames in 5.001 seconds = 296.941 FPS <- cylinder
1501 frames in 5 seconds = 300.2 FPS
1496 frames in 5.003 seconds = 299.021 FPS
1494 frames in 5.001 seconds = 298.74 FPS
1501 frames in 5 seconds = 300.2 FPS
1501 frames in 5.002 seconds = 300.08 FPS
radeonUpdatePageFlipping allow 1 current 0
radeonUpdatePageFlipping allow 1 current 1
1150 frames in 5.013 seconds = 229.404 FPS <-- teapot 
146 frames in 5.017 seconds = 29.1011 FPS
147 frames in 5.027 seconds = 29.2421 FPS
146 frames in 5.011 seconds = 29.1359 FPS
146 frames in 5 seconds = 29.2 FPS
147 frames in 5.029 seconds = 29.2305 FPS
146 frames in 5.018 seconds = 29.0953 FPS
146 frames in 5.003 seconds = 29.1825 FPS

there is a guite a difference here.

Teapot is very slow ;) compared with your results :)

> 
> ipers is running @20-22 fps (~580.000 Poly/sec)

532771.7 (poly.sec) @19 - 20.60 fps ....
(not much slower ...)

and here is isosurf ;) benchmark 
$ ./isosurf 
7179 vertices, 7177 triangles
radeonUpdatePageFlipping allow 1 current 0
Nr unique vertex/normal pairs: 2723
num_tri_verts: 21531
primitive (0x10000): GL_TRIANGLE_STRIP, 
render style (0x100): glVertex, 
enabling normal arrays
new flags (0xa93169): glVertex, GL_TRIANGLE_STRIP, lit, 
radeonUpdatePageFlipping allow 1 current 0
radeon_makeX86Normal3fv/195 CVAL 0 OFFSET 14 VAL 405dece0
radeon_makeX86Normal3fv/196 CVAL 4 OFFSET 20 VAL 405dece4
radeon_makeX86Normal3fv/197 CVAL 8 OFFSET 25 VAL 405dece8
radeon_makeX86Normal3fv done
radeonUpdatePageFlipping allow 1 current 1
Benchmarking...
Result:  triangles/sec: 4.4483e+06  fps: 619.8
Benchmarking...
Result:  triangles/sec: 4.44974e+06  fps: 620
Benchmarking...
Result:  triangles/sec: 4.45172e+06  fps: 620.276
Benchmarking...
Result:  triangles/sec: 4.44974e+06  fps: 620


If my memory serves me right - last time I've tried it was much slower
(well it was almost a have a year ago if not even more ;) T&L ? made
guite an improvement.)
> 
> Desktop resolution is 1280x1024@24 as always ;-)

Same here.
> 
> We need the Mesa (p)thread stuff badly 'cause all numbers are from single 
> theard/process mode (the second CPU was all the time 100% idle).
> 
> Cheers,
>       Dieter
> 
> -- 
> Dieter N?tzel
> Graduate Student, Computer Science
> 
> University of Hamburg
> Department of Computer Science
> @home: [EMAIL PROTECTED]

-- 
Zilvinas Valinskas


-------------------------------------------------------
Sponsored by:
ThinkGeek at http://www.ThinkGeek.com/
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel

Reply via email to