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

Newegg Daily Deals

Closed Thread
 
Thread Tools
Old 05-23-03, 02:23 PM   #169
EMunEeE
Me
 
EMunEeE's Avatar
 
Join Date: May 2003
Location: Raleigh, NC
Posts: 550
Send a message via AIM to EMunEeE
Default

Quote:
Originally posted by CMav
What some people here fail to realize is that the degree of cheating (in this case, the percentages that ATI and Nvidia have gained) does not in any way negate the fact that they both did indeed cheat. It doesn't matter if it was a 1.8% gain or a 24% gain from driver cheats...they're both still cheats.

To me, the only real difference I see is that that the Nvidia crew is MUCH better at writing drivers than the ati folks. I'm not condoning cheating, just pointing out the obvious.
Whoa nelly, it has not been proven that ATi has been cheating...so that point is now mout. Even if it was cheating, it does not show who is better at writing drivers, but which company has more performance flaws to cover up. It can be said that rhe R300/R350s performance are due to excellent hardware design rather than driver cheats. I have no idea what the NV3X performance is due to too, but a good chunk of it is driver cheating.
__________________
EMunXT
ABIT AN8 Ultra (nForce4 Ultra) || AMD Athlon64 X2 3800+ || 2GB OCZ Performance PC3200 DDR SDRAM || ATI All-in-Wonder Radeon X1800 XL 256MB || Creative Soundblaster Audigy 2 || 120GB & 80GB Seagate HDs || Lite-On DVD Drive & DVD RW || Antec P160WF Case || Antec TruePower 2.0 430W PSU || Samsung 204T (20.1") & Hitachi CML174SX (17") LCDs

I'm black, I'm still sexy, my computer is now sexier though.
EMunEeE is offline  
Old 05-23-03, 02:24 PM   #170
Humus
Coder
 
Humus's Avatar
 
Join Date: Aug 2002
Location: Toronto, Canada
Posts: 135
Send a message via ICQ to Humus
Default

Quote:
Originally posted by The Baron
Active cheat detection? Hmmmm... how about the ability to slightly change the CRC by the end-user with a program?

Or just give everyone a freelook version.

Or have multiple camera paths.
Drivers don't look at CRC's. I'm not even sure it's possible for a D3D driver to know which executable is being run. If it was, I think nVidia would have used such a way rather than by interpreting shader calls, splash screen textures etc.
__________________
Visit my site for cool demos
Humus is offline  
Old 05-23-03, 02:27 PM   #171
CyberDork
Registered User
 
Join Date: May 2003
Posts: 1
Default

A 1.8% gain or loss i would say is the natural error in the benchmark which could be 2-3%.
If they got a 1.8% drop in benchmarks after changeing the program could easily be from a change in state of the Operating system and computer. Its not an exact science. between reboots i could potentialy get a 10-20% change in benchmark scores on my system.

A consistant drop of +8% would most likely be from cheating however, being greater than the possible error from the hardware.

If ATI only had a 1.8% drop in their benchmark scores i would just write that off as error and nothing more.

Theres no gaurentee either that 2 machines running side by side will be 100% identical for accurate benchmarking.
CyberDork is offline  
Old 05-23-03, 02:28 PM   #172
Nazgul
Witch-King Of Angmar
 
Join Date: Oct 2002
Posts: 18
Default

Quote:
Originally posted by Badboy_12345
Why is this cheating?? I don't get it
If nvidia programmers are so smart so they found another way to produce THE SAME **** but more efficient then why is it a cheat??
Then they can do the same things in games... why would that be a cheat?
If the the graphics are lowered then its another story...
but if a human cant tell the difference then its not a cheat!
its called optimizing, scraping a god awful 3dmark code and doing a 100000000 times better one
A game's purpose is to render a scene as efficiently as possible while still being correct. If an IHV can come up with a better way to handle a piece of the code that produces the same result, only faster, that's all fine and dandy. Note however, that the replacement shaders from Nvidia do NOT produce the exact same result in 3DMark2K3, so even if this were a valid practice they still f'd up.

Example: If Nvidia or ATI were able to rewrite rendering code for UT2K3 that produced the same output, that'd be fine even for the timedemos. Those timedemos are only supposed to be measuring UT2K3 performance, and as long as the new code gets used in-game as well as in the timedemo, the numbers are still valid.


A synthetic benchmark's purpose however, is to analyze how different cards compare when executing the same code. That's the only way the scores can be truly useful. They cannot operate under the assumption that the IHV's "optimized" code will always be applicable to real-world cirumstances. Neither Nvidia nor ATI can guarantee that they'll be able to rewrite EVERY SINGLE piece of shader code to boost performance by a specified amount. The only way to get a neutral evaluation is for all cards to run the exact same shader code. Currently it's been proven that Nvidia doesn't want to play by those rules, and it's possible that ATI doesn't either(albeit to a much smaller degree), that situation is still being investigated.
Nazgul is offline  
Old 05-23-03, 02:29 PM   #173
curio
R3D Pimpin'
 
curio's Avatar
 
Join Date: May 2003
Posts: 10
Default

Quote:
Originally posted by CyberDork
A 1.8% gain or loss i would say is the natural error in the benchmark which could be 2-3%.
If they got a 1.8% drop in benchmarks after changeing the program could easily be from a change in state of the Operating system and computer. Its not an exact science. between reboots i could potentialy get a 10-20% change in benchmark scores on my system.

A consistant drop of +8% would most likely be from cheating however, being greater than the possible error from the hardware.

If ATI only had a 1.8% drop in their benchmark scores i would just write that off as error and nothing more.

Theres no gaurentee either that 2 machines running side by side will be 100% identical for accurate benchmarking.
Yeah, but that 1.9% comes from a 8% drop in GT4 - which IMO is hardly a margin of error.

But that doesn't even come close to the 50% drop in GT4 that the FX shows...
curio is offline  
Old 05-23-03, 02:32 PM   #174
Miester_V
Apprentice's Master
 
Join Date: May 2003
Location: U.S.A
Posts: 140
Send a message via AIM to Miester_V
Default

Nvidia cheated on 3DMark as well as on all of you FX users. How do you feel? Proud?...You shouldn't.
__________________
OS: Win XP Pro
CPU: AMD XP 2400+
Mobo: Epox 8K7a
PSU: Mortec 300w
Memory: Crucial 2100 512MB DDR
VGA: ATI Radeon 9800 Pro Retail
Hard Drive: Maxtor 40GB 7200rpm
DVD ROM: Pioneer 16X
CD-RW: Plextor 16/10/40
Sound Card: Sound Blaster Audigy
Speakers: Boston Acoustics 7800 4.1

------------------------------------------------
It is my DUTY to not purchase any Microsoft products.
Miester_V is offline  
Old 05-23-03, 02:38 PM   #175
EMunEeE
Me
 
EMunEeE's Avatar
 
Join Date: May 2003
Location: Raleigh, NC
Posts: 550
Send a message via AIM to EMunEeE
Default

Quote:
Originally posted by Miester_V
Nvidia cheated on 3DMark as well as on all of you FX users. How do you feel? Proud?...You shouldn't.
I do feel for the 5800 Users. They were abandoned and cheated on. They should be the furious ones. Cruel World...hate to say...I told you soo...nah just playin...but I think nVidia should just drop the whole GeForce name now that its tarnished.
__________________
EMunXT
ABIT AN8 Ultra (nForce4 Ultra) || AMD Athlon64 X2 3800+ || 2GB OCZ Performance PC3200 DDR SDRAM || ATI All-in-Wonder Radeon X1800 XL 256MB || Creative Soundblaster Audigy 2 || 120GB & 80GB Seagate HDs || Lite-On DVD Drive & DVD RW || Antec P160WF Case || Antec TruePower 2.0 430W PSU || Samsung 204T (20.1") & Hitachi CML174SX (17") LCDs

I'm black, I'm still sexy, my computer is now sexier though.
EMunEeE is offline  
Old 05-23-03, 02:42 PM   #176
Miester_V
Apprentice's Master
 
Join Date: May 2003
Location: U.S.A
Posts: 140
Send a message via AIM to Miester_V
Default

Quote:
Originally posted by $E Mun$
I do feel for the 5800 Users. They were abandoned and cheated on. They should be the furious ones. Cruel World...hate to say...I told you soo...nah just playin...but I think nVidia should just drop the whole GeForce name now that its tarnished.
It's been tarnished from the begining for their horrible I.Q...But that's a whole 'nother discussion...I don't dare to go off topic from this interesting development.
__________________
OS: Win XP Pro
CPU: AMD XP 2400+
Mobo: Epox 8K7a
PSU: Mortec 300w
Memory: Crucial 2100 512MB DDR
VGA: ATI Radeon 9800 Pro Retail
Hard Drive: Maxtor 40GB 7200rpm
DVD ROM: Pioneer 16X
CD-RW: Plextor 16/10/40
Sound Card: Sound Blaster Audigy
Speakers: Boston Acoustics 7800 4.1

------------------------------------------------
It is my DUTY to not purchase any Microsoft products.
Miester_V is offline  

Old 05-23-03, 02:43 PM   #177
DaveM
Registered User
 
Join Date: May 2003
Posts: 82
Default

Shame on nVidia, this makes you that much lower in my book.

I thought the 5900 Ultra would be a redemption for nVidia but now I see that maybe ATI is the definate way to go.

Even if they cheated with a 8 percent in 1 test compared to nvidia in all the test for 24+ percent.

Shame shame.
DaveM is offline  
Old 05-23-03, 02:45 PM   #178
EMunEeE
Me
 
EMunEeE's Avatar
 
Join Date: May 2003
Location: Raleigh, NC
Posts: 550
Send a message via AIM to EMunEeE
Default

All this cheating...I might go back down to my integrated Intel Graphics...at least I know that it sucks due to hardware and not because of cheats..:ROFL:
__________________
EMunXT
ABIT AN8 Ultra (nForce4 Ultra) || AMD Athlon64 X2 3800+ || 2GB OCZ Performance PC3200 DDR SDRAM || ATI All-in-Wonder Radeon X1800 XL 256MB || Creative Soundblaster Audigy 2 || 120GB & 80GB Seagate HDs || Lite-On DVD Drive & DVD RW || Antec P160WF Case || Antec TruePower 2.0 430W PSU || Samsung 204T (20.1") & Hitachi CML174SX (17") LCDs

I'm black, I'm still sexy, my computer is now sexier though.
EMunEeE is offline  
Old 05-23-03, 02:46 PM   #179
hithere
 
Join Date: Jan 2003
Posts: 176
Default

Quote:
Originally posted by The Baron
Jesus Christ, why are we supposed to "trust" a company? A company is designed for one thing--to get YOU to give YOUR MONEY to THEM.
So are muggers and theives.

I would not put blind faith in any company...but I will say that I am not at all certain that the FX in any of its current forms can perform well in any environment without specific code written for it...and that is bad. In fact, it seems to me that Nv30-specific paths are the only things that carry the potential for the card to outperform the 9800. Those paths may be there in Doom 3, will they be there in every game involving shaders, or are you going to have to wait for coders to get around to writing an NV30 path, or for nvidia to get around to writing shaders for every game that comes out.

Last edited by hithere; 05-23-03 at 02:52 PM.
hithere is offline  
Old 05-23-03, 02:49 PM   #180
Blomman00
Registered User
 
Join Date: Mar 2003
Posts: 23
Default

Quote:
Originally posted by Miester_V
Nvidia cheated on 3DMark as well as on all of you FX users. How do you feel? Proud?...You shouldn't.
It's just a card, i don't sleep with it, just play games, and so far have no problems whatsover with my 5800u.
I do lost in faith of Nvidia after this cheating stuff, but as long as my games works i'll keep it.
Blomman00 is offline  
Closed Thread


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 09:35 PM
Current NVIDIA FreeBSD graphics driver releases zander NVIDIA FreeBSD 0 01-27-09 06:22 PM
Current NVIDIA Linux graphics driver releases AaronP NVIDIA Linux 0 11-06-08 05:39 PM

All times are GMT -5. The time now is 07:33 AM.


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