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

Newegg Daily Deals

Reply
 
Thread Tools
Old 05-25-03, 02:49 AM   #49
Behemoth
radeon 9800 pro
 
Behemoth's Avatar
 
Join Date: Apr 2003
Location: Darkness Falls
Posts: 841
Default

the "shuffle instruction" optimization is just like having external help from ati, ati tells the shaders "dont do this dont do that skip this skip that, just listen!! coz we know it doesnt matter on GT4 water and sky". in a way, they are just partially pre-calculated siturational shaders.
otherwise, i cant understand why ati wont use this shuffle instruction optimiztion if it works in every case just like a normal PS2 shader.
Behemoth is offline   Reply With Quote
Old 05-25-03, 03:15 AM   #50
Uttar
Registered User
 
Uttar's Avatar
 
Join Date: Aug 2002
Posts: 1,354
Send a message via AIM to Uttar Send a message via Yahoo to Uttar
Default

Your understanding of shuffle instruction is not correct.

The tradtional ( read: NV20, NV25, NV30, ... ) architectures work on Vec4s. The R300 works on Vec3s and on Scalars at the same time.

This results in improved performance if you can run both of them in parallel.

All they're doing is saying "Do this before that instead of after that" - nothing more. This will result in no IQ difference, and the shader will still work in all cases.

IMO, this is a perfectly valid optimization, and ATI is really only removing it to make sure people who don't know what they're talking about don't spread BS about them cheating.


Uttar
Uttar is offline   Reply With Quote
Old 05-25-03, 03:22 AM   #51
micron
Registered User
 
Join Date: May 2003
Location: U.S.
Posts: 165
Send a message via MSN to micron
Default

Quote:
Originally posted by Uttar
Your understanding of shuffle instruction is not correct.

The tradtional ( read: NV20, NV25, NV30, ... ) architectures work on Vec4s. The R300 works on Vec3s and on Scalars at the same time.

This results in improved performance if you can run both of them in parallel.

All they're doing is saying "Do this before that instead of after that" - nothing more. This will result in no IQ difference, and the shader will still work in all cases.

IMO, this is a perfectly valid optimization, and ATI is really only removing it to make sure people who don't know what they're talking about don't spread BS about them cheating.


Uttar
OMFG I love you Uttar!....that was the most brilliant, perfect explanation of the ATi situation I've ever seen written.
__________________
Athlon64 3000+ "Venice" @2.48Ghz
Asrock Dual Sata2 939
1GB Corsair DDR 400
HIS X1900GT IceQ 3 @700/1600 1.45v
micron is offline   Reply With Quote
Old 05-25-03, 03:43 AM   #52
Behemoth
radeon 9800 pro
 
Behemoth's Avatar
 
Join Date: Apr 2003
Location: Darkness Falls
Posts: 841
Default

Quote:
Originally posted by Uttar
Your understanding of shuffle instruction is not correct.

The tradtional ( read: NV20, NV25, NV30, ... ) architectures work on Vec4s. The R300 works on Vec3s and on Scalars at the same time.

This results in improved performance if you can run both of them in parallel.

All they're doing is saying "Do this before that instead of after that" - nothing more. This will result in no IQ difference, and the shader will still work in all cases.

IMO, this is a perfectly valid optimization, and ATI is really only removing it to make sure people who don't know what they're talking about don't spread BS about them cheating.


Uttar
i dont believe it works without external pre-calculation but if what you said was true, i really think ati is stupid not using 8% more performance boost just because people THINK its a cheat, ati could just prove that it was not a cheat by using it properly without doing driver detection and code alteration in next driver.
to me, removing it does not make it more innocent than proving it to be a valid optimization.
but thanks for you view, i just cant believe it. if its valid why remove it? i just cant get it, sorry.
Behemoth is offline   Reply With Quote
Old 05-25-03, 03:48 AM   #53
Behemoth
radeon 9800 pro
 
Behemoth's Avatar
 
Join Date: Apr 2003
Location: Darkness Falls
Posts: 841
Default

by the way if its not too much a trouble can you tell me what exactly did ati shuffle?
Behemoth is offline   Reply With Quote
Old 05-25-03, 04:57 AM   #54
StealthHawk
Guest
 
Posts: n/a
Default

Quote:
Originally posted by Behemoth
but thanks for you view, i just cant believe it. if its valid why remove it? i just cant get it, sorry.
Is it really that hard for you to believe it? That's pretty cynical, especially why you wouldn't believe nvidia was cheating when the clipping planes came to light.

Why the reversal? You said you didn't want to believe that nvidia was cheating without "hard evidence." But you are quick to believe that ATI is cheating? Without any evidence? And now we know that nvidia was employing a multitude of cheats....ATI has only been implicated in GT4, as far as we know it was only the shader re-ordering.
nvidia was lowering IQ, ATI was not.


As for reasons why they would remove it, isn't that simple?
1) They took a lot of flak over the Quack "cheats." <- Yes, I still think that was a cheat. Anyway, it tarnished their reputation. Removing something questionable shows that ATI has integrity.
2) Implicit comparison with nvidia. ATI owns up to what they did, nvidia won't comment on their cheats or admit they were cheating. This makes ATI look like a more honest company.
  Reply With Quote
Old 05-25-03, 11:18 AM   #55
Sean P.
www.pcper.com
 
Join Date: Mar 2003
Location: Manchester, NH
Posts: 69
Send a message via ICQ to Sean P.
Default

Wow.....I leave for a week and come back to find the whole world turned upside-down...

The events we've witnessed over the past week are certainly a cause for concern. Given the severity of the accusations and implications we have seen, it is far too easy to make a quick judgement on this matter. However, I believe that we should take a step back and wait until we have all the information before we make up our minds entirely. Should the final collection of information prove NVIDIA to be in the wrong, you can be sure that a flame-fest is warranted and justified.

Don't get me wrong...right now I tend to believe that NVIDIA was caught in the act of cheating. However, since we do not know all the details from BOTH sides, this initial impression cannot be solidified into a final opinion. Though I have not had the opportunity to speak with Kyle on this matter, I tend to believe that he feels the same way. As a result, ( I believe ) he is waiting to speak fully on the matter until all the information is given to him.

For one reason or another, people seem quick to tackle Kyle for not jumping on this matter. Though we may or may not agree with his approach on this matter, I think that he deserves the benefit of the doubt here. HardOCP readers come to our website for the cold [H]ard truth...with all the PR bull**** cast to the side. Kyle has run [H] that way from the start and I don't think he is planning on changing that anytime soon...

In the end, let's trust Kyle ( and the rest of the media ) to do the right thing...Wait for all the information before making your mind up and let's see what happens in the coming weeks...

If anyone has any direct questions they would like to ask me, please email me...

pelly@hardocp.com

Have a good one guys...
__________________
Sean "Pelly" Pelletier
Co-Founder / Editor
PC Perspective
www.pcper.com
Sean P. is offline   Reply With Quote
Old 05-25-03, 11:40 AM   #56
DivotMaker
 
Join Date: Jul 2002
Posts: 823
Default

Quote:
Originally posted by Sean P.

Don't get me wrong...right now I tend to believe that NVIDIA was caught in the act of cheating. However, since we do not know all the details from BOTH sides, this initial impression cannot be solidified into a final opinion. Though I have not had the opportunity to speak with Kyle on this matter, I tend to believe that he feels the same way. As a result, ( I believe ) he is waiting to speak fully on the matter until all the information is given to him.

I think you make some reasonable assumptions.

One question that hasn't been asked, or at least I have not seen it asked.

Are the Detonator 44.03 drivers going to be the shipping drivers for FX 5900 U and other 5900 cards? The reason I ask is, nVidia just announced the NV35 and all the "reviews" have been "previews" of potentially shipping product. So far, I don't believe nVidia has provided retail boards/drivers for "review".

I do realize that the 44.03's are WHQL and have been posted for use. The fact that they do what they do in 3D Mark 2003 is well documented. I don't agree with the way the 44.03's or prior FX drivers perform while running the 2003 benchmark.

Could it be that nVidia is just not up to speed with their shader operations/drivers? I don't know, but it seems to be a weakness up to this point in the NV30+ series and I am just wondering if it is indeed hardware-related or driver-related.

Could this also be the beginning of a new API like Glide for 3dfx? Could Cg be the next Glide and is this the reason that nVidia is not up to speed with shader performance in 3D Mark?

I certainly do not have the answers, but I have a hard time accepting nVidia's current position. They are awfully quiet and I don't feel we have heard the whole story yet.

Food for thought...
DivotMaker is offline   Reply With Quote

Old 05-25-03, 12:05 PM   #57
Behemoth
radeon 9800 pro
 
Behemoth's Avatar
 
Join Date: Apr 2003
Location: Darkness Falls
Posts: 841
Default

Quote:
Originally posted by StealthHawk
Is it really that hard for you to believe it? That's pretty cynical, especially why you wouldn't believe nvidia was cheating when the clipping planes came to light.

Why the reversal? You said you didn't want to believe that nvidia was cheating without "hard evidence." But you are quick to believe that ATI is cheating? Without any evidence? And now we know that nvidia was employing a multitude of cheats....ATI has only been implicated in GT4, as far as we know it was only the shader re-ordering.
nvidia was lowering IQ, ATI was not.


As for reasons why they would remove it, isn't that simple?
1) They took a lot of flak over the Quack "cheats." <- Yes, I still think that was a cheat. Anyway, it tarnished their reputation. Removing something questionable shows that ATI has integrity.
2) Implicit comparison with nvidia. ATI owns up to what they did, nvidia won't comment on their cheats or admit they were cheating. This makes ATI look like a more honest company.
yes it is hard to believe.
i didnt say there was no chance nvidia was cheating, i even said in one post that the clipping plane issue was possibly nvidia cheat.
now futuremark caught ati doing driver detection and code alteration, and ati have admitted they did shuffle the instructions, these are hard evidences, they say much more than just a few screenshots and a guess in ET site, nvidia didnt admit anything, fureturemark did not provide any evidences, so i dont think it is cynical, a few corrupted screenshots plus a guess means very different to me than an ati confession plus futuremark evidences.
and yes i didnt believe a few screenshots plus a guess were evident enough to prove nvidia cheating.

i am sorry, i cant really understand how stop doing a suspicious cheat again makes a company more honest than just to prove it was a valid optimization. when people think its a cheat, ati appease them by not doing it again? does that mean ati agreed with people that it was a cheat? or cheat once in a while is ok, forgive-able?

as for gaining speed by shuffling instructions, i can only think they have changed the flow of control that results in skipping part of shader codes and workload that is irrelevant to GT4 sky and water. of coz this is just my guess.

Last edited by Behemoth; 05-25-03 at 12:08 PM.
Behemoth is offline   Reply With Quote
Old 05-25-03, 12:21 PM   #58
Hanners
Elite Bastard
 
Hanners's Avatar
 
Join Date: Jan 2003
Posts: 984
Default

Quote:
Originally posted by Sean P.
Don't get me wrong...right now I tend to believe that NVIDIA was caught in the act of cheating. However, since we do not know all the details from BOTH sides, this initial impression cannot be solidified into a final opinion. Though I have not had the opportunity to speak with Kyle on this matter, I tend to believe that he feels the same way. As a result, ( I believe ) he is waiting to speak fully on the matter until all the information is given to him.

For one reason or another, people seem quick to tackle Kyle for not jumping on this matter. Though we may or may not agree with his approach on this matter, I think that he deserves the benefit of the doubt here. HardOCP readers come to our website for the cold [H]ard truth...with all the PR bull**** cast to the side. Kyle has run [H] that way from the start and I don't think he is planning on changing that anytime soon...
With all due respect, I'm not sure what further evidence you are expecting to see in the next week or two. It's not going to get any more clear-cut than this.
__________________
Owner / Editor-in-Chief - Elite Bastards
Hanners is offline   Reply With Quote
Old 05-25-03, 12:25 PM   #59
Sean P.
www.pcper.com
 
Join Date: Mar 2003
Location: Manchester, NH
Posts: 69
Send a message via ICQ to Sean P.
Default

I think NVIDIA will respond with some sort of technical information to try and prove their case...What we've heard so far is a quick response...I'd imagine a more thorough and formal response will be coming sometime this week...

Once we have this information, we can look at both sides and then make our final decisions...
__________________
Sean "Pelly" Pelletier
Co-Founder / Editor
PC Perspective
www.pcper.com
Sean P. is offline   Reply With Quote
Old 05-25-03, 12:27 PM   #60
Hanners
Elite Bastard
 
Hanners's Avatar
 
Join Date: Jan 2003
Posts: 984
Default

Quote:
Originally posted by Sean P.
I think NVIDIA will respond with some sort of technical information to try and prove their case...What we've heard so far is a quick response...I'd imagine a more thorough and formal response will be coming sometime this week...

Once we have this information, we can look at both sides and then make our final decisions...
I hope you're right, but I get a nasty feeling all we'll be seeing is another attempt to discredit FutureMark, possibly disguised as a technical document.
__________________
Owner / Editor-in-Chief - Elite Bastards
Hanners 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
"GPU has fallen off the bus" error on 650M unless a CUDA program run first amonakov NVIDIA Linux 0 06-18-12 06:34 PM
Getting the proprietary nvidia driver to run with Debian 3.0 r0 (woody) Katchina404 NVIDIA Linux 9 01-12-03 08:49 AM
On The 333 Boards Does Anyone Run At 166fsb john19055 CPUs, Motherboards And Memory 22 07-30-02 07:39 PM

All times are GMT -5. The time now is 12:44 AM.


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