Go Back   nV News Forums > Hardware Forums > Benchmarking And Overclocking

Newegg Daily Deals

Reply
 
Thread Tools
Old 04-15-03, 09:18 AM   #37
SnakeEyes
Registered User
 
SnakeEyes's Avatar
 
Join Date: Jul 2002
Location: Lake Zurich, IL
Posts: 502
Send a message via ICQ to SnakeEyes
Default

That's one definition of cheating. However, a benchmark is developed to show the performance of various different competing pieces of hardware and / or software when doing a certain specific task- ie. rendering a scene as close to a given quality level as possible (in this case, the reference rasterizer image would be really useful). If the hardware is being used in a manner that produces as close to that reference as the hardware is capable of, then I wouldn't consider it cheating. On the other hand, purposely using something less than the hardware is capable of, while knowing that it moves the final result further from the reference, with only the intent of improving the apparent performance of said hardware is cheating in my book. That's why I considered the Quack thing cheating at the time, and why I also consider what nVidia did cheating. (is doing? dunno until someone examines the latest drivers against the reference more closely and finds definite proof that it's slipping something, or else nVidia comes out and says that they are doing something to improve performance at the expense of quality, which btw, is NOT gonna happen).

It's sort of a moot point though, because the issue has been beat to death now. The reality is that nV30 isn't going to remain on the market long, has too many known shortcomings with current software titles (when compared to the competition), and will be replaced in short order. If all the rumors hold true, then nv35 should actually fulfill all the expectations that so many of us had for nv30.

Oh, and before anyone decides to include me in the ATI fanboy camp, the 9700Pro I have now is the first ATI I've had in a loooong time. Prior to this card I've had 5 generations of nVidia cards, and mostly loved every one of them. More than the hardware issues with nv30, nvidia's conduct during this period has turned me off to them. Hopefully they will improve that situation along with the hardware once nv35 finally has had the wraps taken off of it.
__________________
Snake-Eyes
SnakeEyes is offline   Reply With Quote
Old 04-15-03, 09:56 AM   #38
RobHague
I like cheese.
 
RobHague's Avatar
 
Join Date: Mar 2003
Location: UK
Posts: 904
Default

Just a side note that confuses me here....

Even if the 3dmark comes down to 3000-4000 with FP32 and WHQL drivers(just considering for a moment all this is true etc).

Considering the FX is rendering the scenes to a higher precision thats not a bad score... everyone goes off blazing about IQ. Lower score for higher precision? If it was the otherway around with the ATi card you would get people sticking up for it in droves...
__________________

There used to be a signature here, but now there isnt.
RobHague is offline   Reply With Quote
Old 04-15-03, 11:33 AM   #39
gokickrocks
Registered User
 
Join Date: Nov 2002
Posts: 409
Default

dx9 sdk reference rasterizer pic for frame 800
__________________
"never argue with an idiot, they will bring you down to their level, and beat you with experience"
gokickrocks is offline   Reply With Quote
Old 04-15-03, 11:36 AM   #40
gokickrocks
Registered User
 
Join Date: Nov 2002
Posts: 409
Default

the only thing different that i can see (keep in mind that i just woke up to post this), is that the nv30 has a little more brightness, but other than that, the new driver's iq at defaults seems fine to me
__________________
"never argue with an idiot, they will bring you down to their level, and beat you with experience"
gokickrocks is offline   Reply With Quote
Old 04-15-03, 11:41 AM   #41
GlowStick
CoD4!
 
GlowStick's Avatar
 
Join Date: Feb 2003
Location: Florida
Posts: 5,786
Send a message via AIM to GlowStick
Default

I competely agree, the brightness is diffrent, and some textures are a bit sharper. But the raster image dose look better than my 9800pro but barely, the background texture is just not as crisp.

THat raster is a really cool toy heh.
GlowStick is offline   Reply With Quote
Old 04-15-03, 11:42 AM   #42
jbirney
Registered User
 
jbirney's Avatar
 
Join Date: Jul 2002
Posts: 1,430
Default

Quote:
Originally posted by Uttar
Well, you know, I actually think the NV30 is really only FP32 - no FP16 calculations supported.
The only difference is temporary registers who can be either FP16 or FP32, and register usage is very important on the NV3x ( the least, the better )
Thats the point I was trying to make. By sticking to one simple paths things are easier....meaning less transistors and easier optiomizations (remember jc said that scheduling for the NV30 was tricky). I am not saying either is better. Just saying that wiht every engineering choice is a trade offs. they both have "good" and "bad" parts...
jbirney is offline   Reply With Quote
Old 04-15-03, 11:45 AM   #43
RobHague
I like cheese.
 
RobHague's Avatar
 
Join Date: Mar 2003
Location: UK
Posts: 904
Default

Im not sure about the brightness - i think it might have something to do with the fact that its a .jpg (the image of the other drivers). But the 43.51 pic does look sharper (but thats also most likley due to the jpeg compression). But in the issues that matter here is the IQ compared to the R300/350 and it looks fine to me. Can anyone point me too a PNG of the R300/350 on Frame 800 of this test? Id like to look closer you see.

Also maybe we should spend less time comparing and dwelling on which companies product is 'better' and just actually concentrate on the products in question rather than trying to convert people to their product? We have all seen millions of threads starting "Ive used nvidia products for years but...." and there is really NOTHING new to be said on this subject. The thread has started to go off slightley on a R300/350 vs NV30 theme again.

Oh and Fusion love your signature

Quote:
Gamers who don't use AA + AF are morons now ?
Guess the world is full of morons, then ?
I don't PLAY games, I merely look at bitmap pics of games blown up 10x.
__________________

There used to be a signature here, but now there isnt.

Last edited by RobHague; 04-15-03 at 11:51 AM.
RobHague is offline   Reply With Quote
Old 04-15-03, 11:55 AM   #44
GlowStick
CoD4!
 
GlowStick's Avatar
 
Join Date: Feb 2003
Location: Florida
Posts: 5,786
Send a message via AIM to GlowStick
Default

Rob, ill be happy to take a frame PNG from a R350, could someone give me instructions of how to do it?

Update!

Ok figured it out, it was real easy LOL

Here is what the options.

IN the Radeon Control Panel, all options set to Application Prefrence. Therefore no AA is used, and the rest is default.

The Res was 1024x768, i thought that was small, but i belive thats what the other screens were at. If you want larger pictures, i can do bigger, probly 1280x1024 is highist i can do due to my monitor.

THe link is here, warning its like 2.6MB, also it was saved in .bmp format, 3dmark did it by itself


Heh, my ISP has cut off my web hosting bandwith.... heh, if anyone can host it, ill send you a zip E-mail me glowstick@insightbb.com

Last edited by GlowStick; 04-15-03 at 12:05 PM.
GlowStick is offline   Reply With Quote

Old 04-15-03, 12:07 PM   #45
gokickrocks
Registered User
 
Join Date: Nov 2002
Posts: 409
Default

dx9 sdk reference rasterizer (png)

9700 pro(png)

the 9700 pro is pretty much the same as the reference rasterizer

edit:

i did an experiment by converting the pngs into jpgs...and well, the jpg does up the brightness a bit, so the 5800 pic with the new drivers may be correct...

dx9 sdk reference rasterizer (jpg)

9700 pro (jpg)

darkcrow's 5800 w/ 43.51 drivers (jpg)
__________________
"never argue with an idiot, they will bring you down to their level, and beat you with experience"

Last edited by gokickrocks; 04-15-03 at 12:13 PM.
gokickrocks is offline   Reply With Quote
Old 04-15-03, 12:19 PM   #46
GlowStick
CoD4!
 
GlowStick's Avatar
 
Join Date: Feb 2003
Location: Florida
Posts: 5,786
Send a message via AIM to GlowStick
Default

Alright, thx gokickrocks for doing all of that hard work for us!

The images look pirtty good, but the GFFX blurs the background part of the trees. The R300 and Raster arent blured as much.
GlowStick is offline   Reply With Quote
Old 04-15-03, 12:21 PM   #47
gokickrocks
Registered User
 
Join Date: Nov 2002
Posts: 409
Default

i have to agree that the 5800 does blur the background
__________________
"never argue with an idiot, they will bring you down to their level, and beat you with experience"
gokickrocks is offline   Reply With Quote
Old 04-15-03, 01:20 PM   #48
RobHague
I like cheese.
 
RobHague's Avatar
 
Join Date: Mar 2003
Location: UK
Posts: 904
Default

It's possible that the IQ setting was set to balanced or performance though for DarkCrows snapshot.

Tell me what tool i need to be able to take a screenshot of that exact frame and i will post a shot of it set to "Quality" just to make sure.
__________________

There used to be a signature here, but now there isnt.
RobHague is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


Similar Threads
Thread Thread Starter Forum Replies Last Post
Need Help Installing NVIDIA Tesla M2070Q in Linux RHEL5 Ferianto85 NVIDIA Linux 0 05-18-12 08:35 PM
Rumor regarding lack of 680 availability ViN86 Rumor Mill 6 05-09-12 04:48 PM
NVIDIA could rule if they really wanted to Vid_craze NVIDIA GeForce 7, 8, And 9 Series 25 08-16-02 05:24 PM

All times are GMT -5. The time now is 08:08 PM.


Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright 1998 - 2014, nV News.