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

Newegg Daily Deals

Reply
 
Thread Tools
Old 05-18-03, 06:24 PM   #97
muzz
 
muzz's Avatar
 
Join Date: Feb 2003
Posts: 816
Default

Of course folks like to brag about their scores....
They like competition, and most geeks either can't play sports, or are too old ,young or for some other reason unable to do so.
Computer games and benchies give them an outlet to compete.

I still think it is a valid tool to determine performance, as long as the WORKLOAD stays what it should be, and the frames are rendered correctly visually ( what you can see, but NOT at the expense of less workload).

I've spent hours tweaking my system to get high benchies, as I am now too old to play sports ( and I miss them ) competitively.
And I am a competitive person, so I enjoyed the competition a bit....... I need new gear though, so I am not doing any benching these days... thats what happens when you are laid off for awhile, and savings ned to be used elsewhere.... like survival!! heheh.
I see nothing wrong with folks doing what makes them happy, as long as it is not affecting others in a negative way.

I may think folks are strange for some of the stuff they do, but I say whatever floats ya boat.

Edit. man my spelling is atrocious, even moreso than normal.
__________________
muzz
muzz is offline   Reply With Quote
Old 05-18-03, 06:25 PM   #98
Nv40
Agent-Fx
 
Nv40's Avatar
 
Join Date: Aug 2002
Location: everywhere
Posts: 2,216
Default

[quote]Originally posted by legion88
Using image quality as a means to determine whether it is a cheat or an optimization is laughable. It is just as stupid (well, actually, it is almost as stupid) as ATI morons using the totally idiotic notion that if the newer drivers showed similar speed improvements without image quality degradation then that proves the previous drivers weren't cheats.

Cheat means to violate the rules deliberately (http://dictionary.reference.com/search?q=cheat). In the case of FutureMark's products (or any benchmark for that matter), the companies are expected not to "optimize" specifically for their benchmark. That is, it is expected that the optmizations would be apparent for any Direct3D applications, not just 3DMark200x. So those attempts at blurrying the textures in 3DMark2001 (that ATI was also caught doing) on the Radeon 8500s would be obvious examples of cheats.

[/b]

Using image quality as a means to determine whether it is a cheat or an optimization is laughable.
-----------------------------------------------------------------------------
mmmm.. this is very questionable ,because the ATI QUack Fiasco was
because ATI ,cutting corners ,not rendering everything ,decreasing the IQ QUake3 ,just to Inflate their scores ,in that benchmark.

it is questionable too, because is exactly why some B3d members here
says NVidia is cheating in 3dmark.. (dancing pixels) /dull sky and now
the clipping planes. very questionable my friend.

Cheat means to violate the rules deliberately
-------------------------------------------------
and what are the RUles?
Where is says in 3dmark2003 clipping planes/occlusions culling
(what is Valid in games -when the IQ is never lost ),is not valid in the test?
SHow me the RUles ,where it says what Nvidia has done is not valid?
IF FUturemark RUles says that Nvidia and ATI need to do the same things
in the SAME WAY.. then IS FUTUREMARK who is breaking the rules.
because ATI was rendering everything at 96bits and NVidia at 128bits.
IT IS FUTUREMARK which is not HONEST to the PUblic in saying that
the bechmark is and apples vs apples comparisons.

i have never seen in fair contest (apples vs apples) ,where the rules states
that everyone needs to do the same ,,where one runner starts in the middle of the race and the other in the beginning of it.

another interesting question is ..
DOes Future mark rules apply to non_members?
If there are RULES .. does they APPLY to the people WHICH ALREADY STATED IN PUBLIC THAT THEY DONT AGREED?

THE PUBLIC doesnt know how the rules benefits one or the other
in the contest. they only know about the SCORES of each card.
3dmark is clearly .....

not informing well the public - ,gamers ,OEMS ,when using a
benchamrk that make believe people that what they see is apples vs apples comparisons when its not.

not informing well the public - when its make them believe is an acurrate representation of the performance of DIrectx9 games of the future ,when the test use Directx7 and DIrectx8 PS/VS in 90% of the tests.

not informing well the public - when making them believe that the performance in the benchamrk will translate to real games of the FUture.

the right thing to do.. is a big DISCLAIMER with each benchmark saying what each card is doing . what Pixel shaders is being used ,what precision.
and another disclaimer saying..

WARNING:
THIS PROGRAM IS NO WAY PRETEND TO REPRESENT WHAT YOU WILL SEE IN YOUR GAMES ,IS ONLY FOR TESTING PURPOSES.

Last edited by Nv40; 05-18-03 at 06:35 PM.
Nv40 is offline   Reply With Quote
Old 05-18-03, 06:25 PM   #99
gordon151
Registered User
 
Join Date: Mar 2003
Posts: 264
Default

Quote:
Originally posted by jimmyjames123
Well I disagree, I don't think it's so concrete. Also, what Futuremark hopes for and what NVIDIA hopes for are two very different things. NVIDIA wants to run the benchmark as fast and as smooth as possible without corrupting image quality that we actually see.



That is silly. Anyone can read the reviews and see the dramatic improvements in IQ and performance virtually across the board for the FX cards using the newest Detonator FX driver.



If NVIDIA can increase speed without corrupting image quality that we see, I would generally say that is a good thing (and it looks like NVIDIA actually improved image quality with this new driver). From Futuremark's or ATI's perspective, obvious they wouldn't consider it a good thing. Depends on what frame of reference you use. I think the FX Detonator driver is a very good thing for FX owners, and they should be very happy that both image quality and frames per second have improved.
I'll try to take a different approach to this, hopefully I'll get through. 3DMark03 is not a game, it is a benchmark. The object of a benchmark or rather this particular benchmark is to formulate an idea of performance from comparisons. These comparisons are to be executed in the same manner across varying platforms and thus if a set scene is to be rendered it must be rendered in the same manner on each platform.

When you effectively alter what is being rendered then you effectively invalidate the score that is produced. That score in no shape or form can be used for any kind of comparison, with the possible exception of comparisons between cards that are altered in the same manner. In this specific context, when you change the rules you are effectively doing something of which the term for it escapes me at the moment .
gordon151 is offline   Reply With Quote
Old 05-18-03, 06:31 PM   #100
muzz
 
muzz's Avatar
 
Join Date: Feb 2003
Posts: 816
Default

Where does it say clipping planes is not valid?
Are you kidding me? Please tell me you are joking.

The DX9 spec is NOT 128bits, it is 96.. it is a DX9 ( partly) benchmark... it is NOONES fault but NV's that they do 128bit ( which is slower).
__________________
muzz
muzz is offline   Reply With Quote
Old 05-18-03, 06:31 PM   #101
Sazar
Sayonara !!!
 
Sazar's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,297
Default

Quote:
Originally posted by jimmyjames123
Then both NVIDIA and ATI are cheating. Like it or not, both of these companies "optimize" for this benchmark. This makes a perfectly level comparison using a synthetic benchmark very difficult, no doubt about that.
no... not quite...

first off when the nv30 initially gained a huge performance boost they were not rendering @ the required precision level specified in the dx9 spec... ie... fp24 and higher... this has been debated on this and other forums to a degree where most people AGREE such a thing was happening..

secondly... there is no indication from any research done that shows that ati is indeed hacking in this benchmark... they have maintained pretty much the same performance from day one... which means 1 of 2 things to me... either they do not have the capability to optimise like nvidia and gain 2 times the performance from 1 driver release to the next (bear in mind there are some who are harping on the presence of ATi as a beta member of futuremarks 3dmark03 suite) or secondly... ati are staying within spec and hence cannot indeed achieve a higher score in 3dmark03 without more efficient drivers... which is why their incremental increases are so small or non-existent in this benchmark compared to the 100% increase nvidia's drivers have netted it and the rather large % increases we have seen with these drivers..

if the playing field were equal both the companies would benchmark using drivers that followed the dx9 spec and rendered the scenes per the specifications of the benchmark makers..
Sazar is offline   Reply With Quote
Old 05-18-03, 06:43 PM   #102
Nv40
Agent-Fx
 
Nv40's Avatar
 
Join Date: Aug 2002
Location: everywhere
Posts: 2,216
Default

in 3danimations /and Games STATIC planes/backgrounds doesnt need to be rendered a million of times *if they never change *, REndering only what it is in motion IS a very efficient way of saving time in Rendering ,
because the IQ is not lost .doing the opposite is extremely inneficient ,bad animation or programming.

for the people that answer me with "But 3dmark is not a game........."
go to my previous post..

the line between optimizations and cheatings are close , and we can spend *years* here discussing what is a cheat or what is not.. without agreeing in anything. 3dmark is a diferent story because is not a game , and we need to study carefully what the RUles of 3dmark are , and how fair or unfair they are for one card of another. in my opinion if the gamer will not notice a decrease in IQ , between X and Y card , any Hack/optimization/or cheat that doesnt decreae the IQ and give a better experience to the gamer is valid.

Last edited by Nv40; 05-18-03 at 06:55 PM.
Nv40 is offline   Reply With Quote
Old 05-18-03, 06:50 PM   #103
Sazar
Sayonara !!!
 
Sazar's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,297
Default

[quote]Originally posted by Nv40
Quote:
Originally posted by legion88
Using image quality as a means to determine whether it is a cheat or an optimization is laughable. It is just as stupid (well, actually, it is almost as stupid) as ATI morons using the totally idiotic notion that if the newer drivers showed similar speed improvements without image quality degradation then that proves the previous drivers weren't cheats.

Cheat means to violate the rules deliberately (http://dictionary.reference.com/search?q=cheat). In the case of FutureMark's products (or any benchmark for that matter), the companies are expected not to "optimize" specifically for their benchmark. That is, it is expected that the optmizations would be apparent for any Direct3D applications, not just 3DMark200x. So those attempts at blurrying the textures in 3DMark2001 (that ATI was also caught doing) on the Radeon 8500s would be obvious examples of cheats.
Quote:

Using image quality as a means to determine whether it is a cheat or an optimization is laughable.
-----------------------------------------------------------------------------
mmmm.. this is very questionable ,because the ATI QUack Fiasco was
because ATI ,cutting corners ,not rendering everything ,decreasing the IQ QUake3 ,just to Inflate their scores ,in that benchmark.

it is questionable too, because is exactly why some B3d members here
says NVidia is cheating in 3dmark.. (dancing pixels) /dull sky and now
the clipping planes. very questionable my friend.

Cheat means to violate the rules deliberately
-------------------------------------------------
and what are the RUles?
Where is says in 3dmark2003 clipping planes/occlusions culling
(what is Valid in games -when the IQ is never lost ),is not valid in the test?
SHow me the RUles ,where it says what Nvidia has done is not valid?
IF FUturemark RUles says that Nvidia and ATI need to do the same things
in the SAME WAY.. then IS FUTUREMARK who is breaking the rules.
because ATI was rendering everything at 96bits and NVidia at 128bits.
IT IS FUTUREMARK which is not HONEST to the PUblic in saying that
the bechmark is and apples vs apples comparisons.

i have never seen in fair contest (apples vs apples) ,where the rules states
that everyone needs to do the same ,,where one runner starts in the middle of the race and the other in the beginning of it.

another interesting question is ..
DOes Future mark rules apply to non_members?
If there are RULES .. does they APPLY to the people WHICH ALREADY STATED IN PUBLIC THAT THEY DONT AGREED?

THE PUBLIC doesnt know how the rules benefits one or the other
in the contest. they only know about the SCORES of each card.
3dmark is clearly .....

not informing well the public - ,gamers ,OEMS ,when using a
benchamrk that make believe people that what they see is apples vs apples comparisons when its not.

not informing well the public - when its make them believe is an acurrate representation of the performance of DIrectx9 games of the future ,when the test use Directx7 and DIrectx8 PS/VS in 90% of the tests.

not informing well the public - when making them believe that the performance in the benchamrk will translate to real games of the FUture.

the right thing to do.. is a big DISCLAIMER with each benchmark saying what each card is doing . what Pixel shaders is being used ,what precision.
and another disclaimer saying..

WARNING:
THIS PROGRAM IS NO WAY PRETEND TO REPRESENT WHAT YOU WILL SEE IN YOUR GAMES ,IS ONLY FOR TESTING PURPOSES. [/b]
coupla things... firstly the workload in 3dmark03 is uniform... including clipping planes reduces the workload and prevents the benchmark from working as it is designed... if this was not an issue than respected sites such as b3d and ET would not have brought this up and futuremark members such as Worm would not be wasting their time looking into this to determine just what is going on...

secondly... if you bother to read not only futuremark's whitepaper regarding 3dmark03... but also their response to what nvidia alleged instead of basing everything off of what nvidia alleged... you would NOT be making some of the accusations you continue to make v/s 3dmark03... it is a matter of spending 5-10 minutes and perhaps a few more to learn how things work basically... thats all...

also I am all for rendering only what is seen in the sense that this will indeed speed things up

if you have been following the way both nvidia and ATi have implemented features in their hardware you will see that this is EXACTLY what they are doing...

but deliberately clipping sections of a scene off the screen in a benchmark with the camera's POV being 'on a rail' is not a METHOD... if it was than nvidia would be implementing this 'feature' in all games don't you think ? the clipping is a DELIBERATE method of not doing the full load of work... reduced load == higher scores == cheating

full load WITH higher scores while following the dx9 spec (in terms of precision) == NOT cheating... but it is obvious this is not happening...
Sazar is offline   Reply With Quote
Old 05-18-03, 06:56 PM   #104
digitalwanderer
 
digitalwanderer's Avatar
 
Join Date: Jul 2002
Location: Highland, IN USA
Posts: 4,944
Default

Quote:
Originally posted by muzz
The funniest part about this 3DMark issue with Jimmyjames is this....

The image quality went up even in 3DMark03....

So What? The image quality of that Bench is just a means to the end, that bench is NOT REALLY about image quality now is it ( it IS supposed to look nice, and show a bit of tommorrows technology, but NOT the MAIN [<<<<---did you get that]FOCUS.
The MAIN focus of that bench is to show what the card is capable of, render SPEED is one of the MAIN [ooops there it is again.... DOH!] functions of it.
HOW LONG DOES CARD A TAKE COMPARED TO CARD B.. period.

There is alot of $ riding on that stupid benchmark, but it is what it is, which is a tool to figure out if card A can render all these things CORRECTLY ( visually), and HOW LONG does it take to do it... PERIOD.
Like it or not, that is the reason that bench exists, and making something NOT render the EXPECTED WORKLOAD ( that is the benchmarks decision, not the manufacturers) invalidates the scores ( call it whatever ya like).

Think what ya feel ya have to.

I like this summary of it. It's a benchmark, not a game. It ain't for measuring gaming performance, it measures how how long it takes a card to complete a series of renderings, not rendering it all is cheating.
__________________
[SIZE=1][I]"It was very important to us that NVIDIA did not know exactly where to aim. As a result they seem to have over-engineered in some aspects creating a power-hungry monster which is going to be very expensive for them to manufacture. We have a beautifully balanced piece of hardware that beats them on pure performance, cost, scalability, future mobile relevance, etc. That's all because they didn't know what to aim at."
-R.Huddy[/I] [/SIZE]
digitalwanderer is offline   Reply With Quote

Old 05-18-03, 06:57 PM   #105
Sazar
Sayonara !!!
 
Sazar's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,297
Default

Quote:
Originally posted by digitalwanderer
cheating.
there's that word again
Sazar is offline   Reply With Quote
Old 05-18-03, 06:58 PM   #106
muzz
 
muzz's Avatar
 
Join Date: Feb 2003
Posts: 816
Default

Quote:
Originally posted by Nv40
in 3danimations /and Games STATIC planes/backgrounds doesnt need to be rendered a million of times *if they never change *, REndering only what it is in motion IS a very efficient way of saving time in Rendering ,
because the IQ is not lost .doing the opposite is extremely inneficient ,bad animation or programming.

for the people that answer me with "But 3dmark is not a game........."
go to my previous post..

the line between optimizations and cheatings are close , and we can spend *years* here discussing what is a cheat or what is not.. without agreeing in anything. 3dmark is a diferent story because is not a game , and we need to study carefully what the RUles of 3dmark are , and how fair or unfair they are for one card of another. in my opinion if the gamer will not notice a decrease in IQ , between X and Y card , any Hack/optimization/or cheat that doesnt decreae the IQ and give a better experience to the gamer is valid.
OK if you say so..... reduce the workload of a card running the defacto benchmark to make the scores jump artificially......

If you say so.
I'm done with this utter nonsense, as it seems no matter what folks say in regards to this, there are others who just either don't get it, or think everything is just hunk-dory the way things stand.
Sad I tell ya.

I'm out
__________________
muzz
muzz is offline   Reply With Quote
Old 05-18-03, 07:01 PM   #107
jjjayb
Registered User
 
Join Date: Jan 2003
Posts: 101
Default

Quote:
in 3danimations /and Games STATIC planes/backgrounds doesnt need to be rendered a million of times *if they never change *, REndering only what it is in motion IS a very efficient of saving time in Rendering ,doing
You are one hundred percent correct. If you can increase the speeds by not rendering what is not necessary that is a good thing. That is what occlusion culling is all about. This is done on the fly by videocards every day.

BUT, and this is a very big BUT, you need to be able to do that on the fly. If your drivers can tell what does not need to be shown by what the program tells it, that is great. If you have to manually go in and tell the drivers "in this specific benchmark the camera moves here, then here, then here, then stops, don't render anything else". That is cheating.

Once again because some of you just don't seem to get it yet. By preprogramming the drivers to ingore parts of the test because you have studied the test beforehand and tell the drivers exactly where the camera is going to go in a specific benchmark that is CHEATING. There is no other way to put it. It is NOT an optimization.

If I go in and take a test with the answers written on my hand that is considered cheating. This amounts to the same thing.
jjjayb is offline   Reply With Quote
Old 05-18-03, 07:02 PM   #108
digitalwanderer
 
digitalwanderer's Avatar
 
Join Date: Jul 2002
Location: Highland, IN USA
Posts: 4,944
Default

Quote:
Originally posted by Sazar
there's that word again
Sorry, how about "fraudulently inflating their scores for the purpose of deceiving consumers" instead?
__________________
[SIZE=1][I]"It was very important to us that NVIDIA did not know exactly where to aim. As a result they seem to have over-engineered in some aspects creating a power-hungry monster which is going to be very expensive for them to manufacture. We have a beautifully balanced piece of hardware that beats them on pure performance, cost, scalability, future mobile relevance, etc. That's all because they didn't know what to aim at."
-R.Huddy[/I] [/SIZE]
digitalwanderer 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
Bored, impressed, and giddy: Our final thoughts on E3 2012 (with photos) News Archived News Items 0 06-13-12 06:00 AM
Thoughts from console owners on NVIDIA's GEFORCE GRID MikeC Console World 11 05-27-12 08:43 AM
Looking for a good 21"/22" Monitor...any thoughts? Guuts General Hardware 13 09-22-02 11:04 AM
Thoughts on the command line as an interface. lunix Microsoft Windows XP And Vista 10 09-12-02 08:44 PM
GTA Thoughts? Typedef Enum Gaming Central 5 09-03-02 04:51 PM

All times are GMT -5. The time now is 05:27 PM.


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