[{Sniping}Waste]

Originally Posted by tombman

Ok guys, maybe this can help you a bit:

Numbers are from my machine:

FX51 @ 2.5ghz, 9800XT @ default

4xAA/16AF, CTF BOTMATCH: 55fps (NV40 >70fps)

FX51 @ 2.75ghz, 9800XT @ 445/380

4xAA/16AF, CTF BOTMATCH: 58fps
(but Minimum was much higher than with non overclocked machine )

So you see: the posted NV40 results WERE CORRECT !!!!

COnsidering new drivers and beta A2 hardware the NV40 just behaves
like any next generation card @ start.

have a nice day

somthing is not right with the numbers.

here is mine.
AXP 1800+ with 9700PRO dual DVI
Botmatch CTF set to max at 1280x1024 4AA 16AF
CPU 2.2gig GPU 370/330 27.5 FPS
CPU 1.6gig GPU 370/330 19.4 FPS
CPU 2.2gig GPU 300/290 25.7 FPS

This game is more limited by the CPU and not the GPU.

[T2k]

Actually 1280x768 is a widescreen 1024x768... that's the next res to 1920x1200 I did setup on my machine... (I'm too busy to do the refresh fix for 1600x1000 in Rage3D Tweak... )

[anaqer]

Surely you didn't mean to compare the scaling of an FX51 over 2.5GHz with that of an AXP below 2.2GHz...

[T2k]

Originally Posted by anaqer

Surely you didn't mean to compare the scaling of an FX51 over 2.5GHz with that of an AXP below 2.2GHz...

1. FYI: AXP on 2.5GHz, as I wrote it
2. that's my poitn: the comparison

ps; errr... who are you talking to, actually? )

[tombman]

Originally Posted by {Sniping}Waste

somthing is not right with the numbers.

here is mine.
AXP 1800+ with 9700PRO dual DVI
Botmatch CTF set to max at 1280x1024 4AA 16AF
CPU 2.2gig GPU 370/330 27.5 FPS
CPU 1.6gig GPU 370/330 19.4 FPS
CPU 2.2gig GPU 300/290 25.7 FPS

This game is more limited by the CPU and not the GPU.

My scores are at 1600x1200 not 1280x1024 like yours. btw, your scores are incredibly low, there is something wrong with your machine anyway.

Furthermore i don´t know what map/bots you benched....

[anaqer]

Originally Posted by T2k

ps; errr... who are you talking to, actually? )

{Sniping}Waste - I thought I was fast enough to comment right after him, I guess I was rong though.

EDIT : I guess this explains the AXP / FX51 part too.

[{Sniping}Waste]

The map is convoy botmatch. I know its not a AFX but it should show if its limited by the CPU or GPU. The test is set to the highist settings on Cat 3.10.
I have a 17" monitor so my max is 1280x1024.

[Fred da Roza]

Originally Posted by Chalnoth

My previous comment around that time, where I stated essentially that I didn't consider the benefits of the R200 over the NV2x worth buying a R200 over, were based on the fact that to date, zero games had made use of the higher precision offered by the R200. We all knew that DOOM3 would, but that was still a while away. Even today, very few games make use of the added precision available on the R200..

And zero games to date have illustrated any benefits with FP32. Hasn't stopped you from claiming FP24 isn't enough for the R300 generation, endlessly complaining about it and claiming it's essential in this upcoming generation of hardware.

Originally Posted by Chalnoth

Please note that this argument (what I will buy) is different, in my mind, from the argument about what I consider to be good for the future of 3D games, and what to be bad for the future of 3D games.

Funny how the higher precision of the R200 (where it could be seen) was not a reason for you to buy it over a GeForce 3. But the higher precision of the NV30 (where it isn't even noticeable) is a major reason to buy it over the R300.

[T2k]

Originally Posted by Fred da Roza

And zero games to date have illustrated any benefits with FP32. Hasn't stopped you from claiming FP24 isn't enough for the R300 generation, endlessly complaining about it and claiming it's essential in this upcoming generation of hardware.

Funny how the higher precision of the R200 (where it could be seen) was not a reason for you to buy it over a GeForce 3. But the higher precision of the NV30 (where it isn't even noticeable) is a major reason to buy it over the R300.

And so on with his other stances like praising NV30 for its "flexible architecture"... ROTFLMAO definitely really good for the future of 3D games, isn't it?

edit: better wording

[Sazar]

meh... chalnoth has his opinion... why does he need to be put through the wringer this whole thread based on it?

[Heathen]

meh... chalnoth has his opinion... why does he need to be put through the wringer this whole thread based on it?

Usually because he states it as absolute fact.

[digitalwanderer]

Originally Posted by Sazar

meh... chalnoth has his opinion... why does he need to be put through the wringer this whole thread based on it?

No real reason, it's just so fun! :P

[Sazar]

yah but this discussion keeps getting side-tracked for x number of reasons not least of all which is personal attacks on the integrity of whomsoever it may be...

I have no doubt we all know chalnoths opinions but its tiresome when you are reading through a 20-30 page thread and 60% of the posts are b.s.

disclaimer: no offense to anyone...

[Evanescence]

NV40 has only EXTRALARGE 4 PIPES

http://home.earthlink.net/~mcwiggin/...ipesLargeb.jpg

[John Reynolds]

Chalnoth's bias is blatantly obvious but at least he doesn't allow himself to be pulled into sniping at other forum members.

[T2k]

Originally Posted by Evanescence

NV40 has only EXTRALARGE 4 PIPES

http://home.earthlink.net/~mcwiggin/...ipesLargeb.jpg

ROTFLMAO!

[Vegetto-eX]

OMG why the 16x1 pipelines is still being discussed? IT is .... really it is 16x1 and for real :?

[digitalwanderer]

Originally Posted by Vegetto-eX

OMG why the 16x1 pipelines is still being discussed? IT is .... really it is 16x1 and for real :?

Shhhhhhh, it's much more fun to watch 'em argue!

[anaqer]

Originally Posted by Vegetto-eX

OMG why the 16x1 pipelines is still being discussed? IT is .... really it is 16x1 and for real :?

Dunno... maybe cause there's no hard EVIDENCE yet?

[StealthHawk]

This UT2004 benchmark arguing is the most ridiculous thing I've ever seen. I thought we had all already agreed several pages back that botmatches at 1600X1200 were not purely CPU limited...

[Neeyik]

Originally Posted by StealthHawk

This UT2004 benchmark arguing is the most ridiculous thing I've ever seen. I thought we had all already agreed several pages back that botmatches at 1600X1200 were not purely CPU limited...

Indeed - even 8 bots in UT2003 isn't CPU limited with a 5950 Ultra:

http://www.beyond3d.com/previews/nvi...index.php?p=11

[Vegetto-eX]

Originally Posted by anaqer

Dunno... maybe cause there's no hard EVIDENCE yet?

What about hearing that from the most trusted sources that we know? even from nv? I repeat it is 16x1

[Dave Baumann]

Nick, I would suggest that UT2003 is considerably less demanding on the CPU front than UT2004 is :

http://www.beyond3d.com/reviews/ati/.../index.php?p=8

(and look at the page before).

The 3DCenter tests are CPU / system bound on the Gericom w 3.06GHz P4, and this is relatively low end graphics.

[Neeyik]

I've not looked at the 3D Center results Dave but yours (sorry for not reading the review earlier...oopsie) suggests to me that, while it is more CPU heavy that 2003, UT2k3 is still not CPU bound at 1600x1200 - which is what (I think) was the main point of the argument. I know I didn't make it clear but this is what I'm on about at least. Once my copy of UT2k4 arrives in the post, I'll have a proper play.

[T2k]

Originally Posted by Vegetto-eX

What about hearing that from the most trusted sources that we know? even from nv? I repeat it is 16x1

Link, pls. Otherwise it's the same crappy rumour like The Inq.