rec.autos.simulators

GPL K6/2 vs K6/3 results.

Greg Campbel

GPL K6/2 vs K6/3 results.

by Greg Campbel » Sun, 02 Jan 2000 04:00:00

The "Which K6" question seems to get asked every few weeks.  Since I
recently spluged on a K6-3/450, here are the results of a bit of testing
I recently did.

GPL 1.2 running at 800x600 on a 90MHz, 12M Voodoo2.  Tyan 1590, 128M, 1M
L2.  All graphics options maxed (ooops..except for mirrors = cars and
track), 16 sounds issuing from a SB-Live. 19 AI cars.   Framerate
numbers are taken from the back the grid, using default track.ini
positioning.  The first framereate is prior to the flag drop, the second
is 2~3 seconds after the "launch," as the field accelerates.

K6-2/450

Monza        9.7 on the grid, 8.x underway.
Spa            8.1, 7.x
Glen            9.3, 8.x

K6-3/450

Monza        13.7, 12.x
Spa            11.5, 10.x
Glen            13.3, 11.x

Although it's difficult to objectively measure, I get the impression
that this ~40% framerate increase continues throughout the race.  The
K6-3 chip produces noticeably smoother gameplay, and keeps the framerate
counter pegged more often.

If I get the time, I'll do some more testing, probably with a smaller AI
field.

FWIW, I also did some Unreal Tournament testing.  Framerate taken at a
fixed spot went from 31.3 to 40.2.

-Greg

--

Note!
Anti-spam tactics engaged!
Remove the duplicate "g" and "s" before replying.

Kraku

GPL K6/2 vs K6/3 results.

by Kraku » Mon, 03 Jan 2000 04:00:00

That's running at 100 MHz FSB, I presume?

Very interesting, thanks for the info.

Kraku

GPL K6/2 vs K6/3 results.

by Kraku » Mon, 03 Jan 2000 04:00:00

Actually, I'm also curious as to how it would compare with my non-CTX K6-2
400, although I guess your K6-2 450 does have the new core?
Jalo

GPL K6/2 vs K6/3 results.

by Jalo » Mon, 03 Jan 2000 04:00:00

Very interesting stuff, I didn't think the faster L2 cache would matter in
3D games. I guess I was wrong.
So I guess looking at the numbers, even a K6-2 500 would be slower than a
K6-3 450?
I guess they'll be even more of these K6 questions when the K6-2+ (doesn't
have a name yet) comes out.

M Chapma

GPL K6/2 vs K6/3 results.

by M Chapma » Mon, 03 Jan 2000 04:00:00

thanks for the info Greg, just what I was looking for. I have a K6-2 350 and
was unsure whether to buy a k6-2 or k6-3 to upgrade.
Joel Willstei

GPL K6/2 vs K6/3 results.

by Joel Willstei » Mon, 03 Jan 2000 04:00:00


    I upgraded from a K6-2/350 this past summer to a K6-2/450 rather than a
K6-3 simply do to availability. But by this summer I hope that the
K7-1000mhz chips are out. That will be my next upgrade along with a Voodo5
(?)

    The only problem that I can see popping up is that I'm presently basing
everything on the requirements of GPL. But what if the new WCS and or
Trans-Am sims takes us by surprise. Will they be optimized for glide ?

Joel Willstein

Jalo

GPL K6/2 vs K6/3 results.

by Jalo » Tue, 04 Jan 2000 04:00:00

You really don't need a Athlon (K7) 1 gig to run GPL, I have a Athlon 550,
and it runs GPL about as well as it can be ran. You really don't need a
Voodoo 5 to run GPL either. But I see what your saying, you just never know
what the next-gen sims are going to require. That WCS looks like it is going
to want a FAST cpu.

Joel Willstei

GPL K6/2 vs K6/3 results.

by Joel Willstei » Tue, 04 Jan 2000 04:00:00


   Glad you got my point.

    But trying to stay current is half of the fun for me.

Joel Willstein

Matthias Flat

GPL K6/2 vs K6/3 results.

by Matthias Flat » Tue, 04 Jan 2000 04:00:00

On Sat, 01 Jan 2000 15:52:39 -0800, Greg Campbell


>GPL 1.2 running at 800x600 on a 90MHz, 12M Voodoo2.  Tyan 1590, 128M, 1M
>L2.  All graphics options maxed (ooops..except for mirrors = cars and
>track), 16 sounds issuing from a SB-Live. 19 AI cars.   Framerate
>numbers are taken from the back the grid, using default track.ini
>positioning.  The first framereate is prior to the flag drop, the second
>is 2~3 seconds after the "launch," as the field accelerates.
>K6-2/450
>Monza        9.7 on the grid, 8.x underway.
>K6-3/450
>Monza        13.7, 12.x
>Although it's difficult to objectively measure, I get the impression
>that this ~40% framerate increase continues throughout the race.  The
>K6-3 chip produces noticeably smoother gameplay, and keeps the framerate
>counter pegged more often.

Thanks, that's quite interesting.

I have a similar system, but with a Rendition Verite 2200 and a K6-2
400 CXT.
I hope I can compare the numbers on the weekend.

--
Matthias Flatt

Greg Campbel

GPL K6/2 vs K6/3 results.

by Greg Campbel » Wed, 05 Jan 2000 04:00:00


> That's running at 100 MHz FSB, I presume?

Yes.

-Greg

--

Note!
Anti-spam tactics engaged!
Remove the duplicate "g" and "s" before replying.


rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.