More 'Application-Specific' Optimizations in NVidia Drivers

EconolineCrush writes "Futuremark and NVIDIA have been embroiled in a spat over various cheat/optimizations in 3DMark03 for several weeks now. Last week, the soap opera appeared to be over; Futuremark and NVIDIA released a joint statement in which Futuremark clarified that NVIDIA was optimizing its drivers for 3DMark03 rather than cheating. This story, however, appears to be far from over. Tech Report has uncovered a new series of optimizations in NVIDIA's Detonator FX drivers that affect image quality in even Futuremark's latest 3DMark03 build. What's more, if you rename the 3DMark03 executable, the optimizations disappear."

Comments

and they all told me I was crazy....

[sweeney37]

well luckily for me I've renamed all my executables files 3DMark03.exe for some time now.

Mike

Re:and they all told me I was crazy....

[Anonymous Coward]

You've been running the NVidia Dawn demo, haven't you?

Re:and they all told me I was crazy....

[orkysoft]

So, do the other people in your network FPS games already accuse you of "wallhacking" ?

Re:and they all told me I was crazy....

[Anonymous Coward]

actually being modded funny doesn't acount for any karma points.

from the faq:
Note that being moderated Funny doesn't help your karma. You have to be smart, not just a smart-ass.

Re:and they all told me I was crazy....

[Anonymous Coward]

That was the most predictable post ever.

Re:and they all told me I was crazy....

[Mysticalfruit]

So NVIDIA is guilty of using corked drivers?

Geez! On a side note, I had heard of something like that where some site recommended that renaming your executable to some "name" would give better results.

I think someone should run "strings" on the nvidia driver to see what other names work...

Re:and they all told me I was crazy....

[Anonymous Coward]

you must be near an enemy altar

So What? Who Cares?

[tomstdenis]

Ooh ooh I scored a 19341 on my 3dMark test so that means I can play Quake now?

Aren't we smart enough not to be pulled in my marketing hyperzor?

Tom

It's the reviewers' fault

[91degrees]

They always use the 3DMark results as though it's some sort of holy scripture, and as though a benchmark can indicate how well it will work in a real everyday situation. Every industry optimises for benchmarks. From a marketing point of view, it's insane not to.

The only reliable way to test is by testing it withthe applications it's used for. Get some actual games, and see what the frame rate is. If they optimise for those tests then it doesn't matter! It means they're oiptmised for real world situations.

Re:It's the reviewers' fault

[curtisk]

Hot damn! You nailed it right on the head!

Sure 3dMark is OK for a rough idea of your graphics throughput, but like you said some take it wayyyyy to seriously. Unfortunately, the testing methods you propose (albeit rational) aren't quite sexy enough to sell cards. And certainly unquotable for the magazine ads

"My solitaire looks so colorful." - Mavis Jones AARP

I know you meant 3D apps, but these cards are used for 2D as well and seems to get overlooked in that arena at times.

Re:It's the reviewers' fault

[Anonymous Coward]

The problem is that there are no real world applications yet that use the DX9 features the latest batch of graphic cards are offering. On current "real world" applications, the speed gain by those cards over the generation before them is not really that big....

Re:It's the reviewers' fault

[Anonymous Coward]

Except that they can "optimize" those timedemos (which means you will get performance boost only in timedemos, not in actual gameplay) similar way as it was done in 3dmark03, you might wanna read last discussion about this.

Re:It's the reviewers' fault

[10Ghz]

You are obviously ignorant. Among others, Dell use 3DMark when the determine what 3D-acceletators to buy. So cheating in it can cause financial damage/gain.

As to the "Just use games to benchmark!". It's not that easy. 3DMark is meant to test vid-cards on demos that use future technologies. Games obviously can't do that, since in order to have reliable benchmarks with them, the games need to be released first. And fact is that games are lagging when it comes to implementing new tech. that's why we need benchmarks like 3DMark, that test those features that are not yet used in games.

you and your like say "Who cares? It's the games that matter". But I think that cheating (no matter what's the app) tells quite alot of the company in question. The fact that NV has been found to cheat (repeatedly) tells me that they are scum

Re:It's the reviewers' fault

[DrXym]

By all means, hardware vendors should use benchmarks and games to optimise performance. A well designed benchmark will exercise a driver in a controlled manner in a way it will be typically in real life and thus is a good way to improve real world performance.

But this is not optimising, it's deliberately cheating. That's what it's called when you ignore the settings you were told to use and substitute in your own faster ones, simply because you know you're running a benchmark programme used by consumers and reviewers to determine performance. Cheating may not a strong enough word - this almost amounts to fraud, claiming one thing and delivering another.

Re:It's the reviewers' fault

[Spellbinder]

i am not such a nvidia friend
but they have the best linux support
except my nforce2 bord has no agp driver for no nvidia cards =((((
and ati has cheated too
there where drivers that got more fps in quake3 and reduced the image quality without telling something to the users/benchmarkers
it is about time that someone writes a good benchmark tool for linux
some part could be synthetically but it would be nice to include some games/demos

Re:It's the reviewers' fault

[91degrees]

Dell use 3DMark when the determine what 3D-acceletators

Yes, Dell shouldn't rely on these benchmarks either. Anyone who makes buying decisions based solely on generic benchmarks is a fool.

As to the "Just use games to benchmark!". It's not that easy. 3DMark is meant to test vid-cards on demos that use future technologies

Like what? The graphics card industry is a mature market now. Features aren't changing. We're just seeing more speed. About the only new feature recently has been programmable shaders, and even those can be tested adequately using DX7 style texture stages.

Actual applications will be written to work around potential inefficiencies anyway. If a game performs badly in a 3DMark test because of latency issues, it's quite possible that the actual games will simply use a slightly modified version of the same shader with different instruction order.

And I do think it's disreputable that NV cheat. However, I don't start by assuming they're a reputable company. This is the behavious I expect from them.

Re:It's the reviewers' fault

[johndiii]

So, a truthful review would say "It does not really matter which card you buy at this time, because most games do not take advantage of the high-end features. By the time that games actually utilize the advanced capabilities of this generation of graphics cards, there will be a new generation, with a new set of (unused) features."

Re:It's the reviewers' fault

[IPFreely]

You would think the easiest way to eliminate this type of cheating is to make it a standard practice to RENAME THE PROGRAM before running any type of benchmark, game, utility, application or whatever. Once you do that, you are running just like any other applicaton, and you see what the drivers really do.

Testing sites should make it standard practice to rename the EXE to some random name before ANY testing is done, and declare such up front on every test just so the manufacturers know it.

It won't prevent all cheating. Manufacturers will find other ways to identify to source app (data profiling, other EXE file signatures) but those are harder to do, take more time to write, and to execute.

Up the anti and see who plays fair and who cheats harder.

Re:It's the reviewers' fault

[DrXym]

Unfortunately renaming is only a short term solution. It would be trivial for the next version of the driver to scan memory for certain bytes unique to the software or hashing some bytes from the executable file, or see what other DLLs it's loading or any manner of things and come to same conclusion.

You'd end up in an arms race, the likes of which anti-virus and anti-spyware find themselves in today. You put in one measure and the opposition counters with something different.

The only solution is for the driver writers to stop cheating in the first place. Perhaps there should be an industry sanctioned penalty for cheats, such as a product review blackout for a month, deducted points, 'cheating bastards' awards etc.

Re:It's the reviewers' fault

[jdew]

nah, matrox has nvidia beat for drivers ^_^. who wants to dork with their kernel at runtime with modules? ick

Re:It's the reviewers' fault

[shaka999]

So I'm supposed to go out and buy a bunch of differenct cards, test my games, and return the ones I don't want?

Consumers need a reliable way to know which cards to purchase. It may be that each tech site out there should come up with their own, original, game demos so many of these optimizations can't be done. But the fact still remains, consumers need a reliable way to know what to buy.

Re:It's the reviewers' fault

[ghztew]

You do have a good point 91degrees, but you are forgetting what one of the major purposes of 3DMark03 is... to determine how these cards will function in upcoming games. i.e. Half-Life 2. Their benchmarking suite pushes DX9 capabilities and at this time, there are not games out there that are taking full advantage of DX9.

Re:It's the reviewers' fault

[S.Lemmon]

One consistent truth I seen in anything technology related over the years is *never* buy something because it will be useful "in the future". Inevitably by the time the "future" rolls around those same features will be a lot cheaper to purchase. To make things worse, the actual requirements often change in unexpected ways making your fancy card obsolete even before the special features become popular. In the end, all you get for being an early adopter is some minor bragging rights and a much thinner pocketbook.

Re:It's the reviewers' fault

[Steveftoth]

Anyone buying a video card for the next generation of games is crazy! Buy the cheapest video card that will let you play the games you can buy TODAY. If and when Doom3, HL2, DNF (ha) come out, then go and buy the card with your goddly perforance then, but not now.

Video cards happen to be ahead of the curve right now. Hopefully the software will catch up and be ahead soon.

Re:It's the reviewers' fault

[afidel]

But some people like to look ahead a few months to what will be coming out. Doom3 and some other titles that will be out by the end of the summer or by xmas shopping season WILL use DX9 (or in the case of Doom3 OpenGL 2.0) features. So if you are building a box now or looking for an upgrade you might very well want to know who has the best DX9 performance so that you won't be underpowered by the fall.

Re:It's the reviewers' fault

[afidel]

Boy are you ignorant, you think you can test a programmable shaders performance using DX7 style shader calls???? No way, DX7 fixed functions are a small subset of the capabilities of a DX9 card, and depending on how the DX7 level features are implemented (fixed paths vs translation to programmable shader code) may not in any way tell you anything about the programmable shader unit. What Nvidia did was not simply reordering instructions, they took a shader that was meant to run at a minimum of FP24 and replaced it with one running at INT12, a huge decrease in quality and a huge illicit gain in performance from a benchmarking perspective.

Re:It's the reviewers' fault

[Mooncaller]

This benchmark is an aid in evaluating the performance of a card with tomorows games. The goal of this benchmark is to stay ahead of the current crop of games. Right now, 3DMark evaluates sytheticaly DirectX 9.0(?) performance. Few games use all of the DX9 spec. but they will. Given this, 3DMark has a valid use, though like any sythetic test, its results need to be taken cum grano salis.

Re:It's the reviewers' fault

[realdpk]

Perhaps benchmarks should start recording every n-th frame (randomly), and then write some program which attempts to measure the % of differences between it and what it "should" render as, to catch all the errors.

I don't know how best to generate the "should" images. But I'd bet someone a lot smarter than me could figure it out, heh.

Re:It's the reviewers' fault

[johndiii]

I wouldn't frown on looking ahead, but I look at it in a tactical sense. If I'm buying hardware against what I'll need six months down the road, I'm going to wait until that time to evaluate how well it performs with the actual software. The price will almost certainly be lower, the drivers will have fewer bugs, and hardware "gotchas" will be out in the open.

The flip side of the competition in benchmarks is price competition. In the Christmas buying season, I would expect to see some price cuts by one or both of these companies.

My buying habits are somewhat different from those that drive the video card wars. I have never bought a bleeding-edge video card, mostly because they just seem so overpriced. And, after a couple of heavily hyped stinkers, I never buy a game without reading a review (and verifying that it will run well on my hardware). Fortunately, with 'Net reviews, the lag time is much less than, for instance, when I was looking at TIE Fighter.

Finally, leading edge performance is not that important to me. Even with a next-to-best card, small changes to the quality settings, and decreasing the resolution, will let me run at a much more than adequate frame rate. And it doesn't decrease the quality of the game.

Re:It's the reviewers' fault

[johndiii]

I've saved a lot of money not buying bleeding edge cards. Right now, I'm really enjoying my Radeon 8500DV. I've been playing with video and such, and I love the remote control that comes with it. I've run video cables all over the place, and I can sit in the living room and control the computer in the den. Even nicer than a Tivo, IMO. I've about decided to rewire the whole house with video (inline amps a must), network cables, and audio. :-)

Re:It's the reviewers' fault

[afidel]

Well when I bought my latest card it was because the new game I was about to buy did not work with my video card (NWN doesn't like 3dfx cards so I had to upgrade from my v3 3k). I looked at what would run NWN well for around $100, most of the cards in that pricerange would run it fine, but then I looked at the benchmarks and noticed that the geforce3 Ti200 was running much better than the Geforce4 MX cards that were around the same pricepoint, did some research and found out about the lack of a shader unit in the MX cards and decided on the Ti200. It turns out that the difference in those cards which would have had little difference in performance for my game at the time (NWN) means that I may not have to buy a new card this fall because Carmack has said that my card is kind of the baseline he is using for Doom3. So by spending a couple dollars more and looking at a benchmark ~18 months before Doom3 came out I was able to save another generation of videocard upgrades.

Re:It's the reviewers' fault

[johndiii]

As was said elsewhere, by the time DNF ships, the current bleeding edge cards will be available for $25 or so.

The reality of the video card market is that there is a lot of money chasing the latest and greatest models. That's the only thing that allows them to charge the prices that they do. Just glad it's not my money that's flying out. :-)

Re:It's the reviewers' fault

[johndiii]

I can't quarrel with your tactics in that instance. Sounds like you got it just right. My last Voodoo card is still in an old computer.

My impression is that the majority of the money chasing the bleeding-edge cards is from much less thoughtful and intelligent buyers, though.

Re:It's the reviewers' fault

[Steveftoth]

That's why they invented Fry's!

Return anything there for no reason at all.

Re:It's the reviewers' fault

[RichardX]

By the time DNF ships civilisation will have long since crumbled, the human race will be a long forgotten memory, the giant ants will have been and gone, and our sun will be getting ready to go supernova...

Re:It's the reviewers' fault

[KU_Fletch]

This is why I miss PC Accelerator. Not only were the reviewers complete assholes who weren't afraid to call a crap game crap, but they used realistic benchmarks on video cards. They'd pick 3 recent games, use the same system and just replace the card. That way you could see that one card might be better are certain games but not others.

Oh well, I hope this will at least create some sort of call for alternative testing measures in the future.

Re:It's the reviewers' fault

[Anonymous Coward]

Sort of...the benchmarks definitely don't matter in a real world setting at all, but I wouldn't agree that testing the frame rate in real games matters much either. After all, who cares if you only get 110 fps, when so and so competing card gets *gasp* 118!

I know when I go to buy a video card, I look for things like driver stability, not speed. But then, I don't like getting into dick contests just for the hell of it.

Re:It's the reviewers' fault

[91degrees]

Boy are you ignorant, you think you can test a programmable shaders performance using DX7 style shader calls

Of course you can! How do you think the cards run legacy apps? Do you think they have a completely separate pipeline to handle the fixed function? Of course they don't. That would be an insane waste of silicon.

All the first shaders gave us was a couple of different ways of addressing texture coordinates. Later version added dependent texture reads. I'll grant that dependent texture reads may affect performance, but generally the bottleneck is going to be the time it takes to read the data, instruction throughput and instruction latency. We get exactly the same bottlenecks when using the fixed function pipeline.

What Nvidia did was not simply reordering instructions, they took a shader that was meant to run at a minimum of FP24 and replaced it with one running at INT12, a huge decrease in quality and a huge illicit gain in performance from a benchmarking perspective.

If it was such a huge loss in quality then why didn't the software detect this? If the benchmark only tests for speed and completely ignores image quality then it's a crap benchmark, and anyone who uses it has no justification to complain.

I didn't say NVidia were reordering instructions. My point was that if the instructions were ordered differently in the benchmark, then we could see some dramatically different results depending on the card. The benchmarks simply do not adequately test the performance of a card. Real world applications don't either, but at least they test the performance on real world applications.

Re:It's the reviewers' fault

[CoreDump01]

it is about time that someone writes a good benchmark tool for linux

Whats wrong with glxgears? ;-)

Re:It's the reviewers' fault

[cait56]

Two Points

• The name of the executable having an effect on the behavior of the driver is a smoking gun. What's the defense, "Oh, that's an optimization interface that we forgot to document?"
• It is the reviewer's fault, too, for being stupid. Good restaurant critics make it a point to dine anonymously so that they will receive the same meal and service as a regular customer would. A test program that does not conceal its identity is asking to be tricked.

Re:It's the reviewers' fault

[doug363]

That's very difficult to do. Different graphics cards use slightly different algorithms to render scenes anyway, so there's no "canonical" or "correct" rendering. Many graphics cards also allow the user to enable or disable options such as antialiasing etc.

Even if you did have a correct rendering and tried to look for scenes that look different from the correct version, it's difficult to quantify how much the differences affect the user's perceived quality of the image. And how do you tell that a slight visual degradation is due to a driver cheat rather than a lower-grade graphics card?

Re:It's the reviewers' fault

[Shardis]

So let's look at it this way... It doesn't matter to you that this performance gain comes into play only if a certain application is used? You think companies should be able to say, "okay, this runs at 90fps", when in fact it only runs at 90fps for *one* application? I know I'm waaaay stretching the technical difference here, but it still matters.

Re:It's the reviewers' fault

[91degrees]

Oh, sure. It's fairly dishonest of the card manufacturers to optimise for a specific benchmark in a way that only comes into play for that benchmark. I just don't think anyone should be surprised that they do this. Especially with the amount of faith vested in the benchmarks be reviewers.

The only solution is to stop having so much faith in the benchmarks. Manufacturers have always optimised for them.

Re:It's the reviewers' fault

[Timberbarn]

The interesting thing about Futuremark is that it is being used as a best guess about how a particular graphics card will run a future, non-existent game. The current controversy involves DirectX version 9 games of which none curently exist. Reviewers have all done us a terrible disservice in using this benchmark. It is not that there is anything particularly wrong with the benchmark, but rather it is a waste of effort. Serious gamers should be interested in how the hardware/software works with the games that they play now or in the very near future. Since no DirectX version 9 game now exist, and since it will be probably more than a year until more than a few such game exist, then any wise investor would wait until such games, that they are interested in playing, exist befor purchasing. By the time the game is available, the current cards will be considered yesterday's technology with lower prices and more stable drivers. The fastest thing will then be newer hardware that you can not reasonable purchase today. Dump Futuremark, who cares

the lengths people will go to...

[Machine9]

...to profit off of their customers hopes and dreams. This kind of crap is really showing off the sorry state of our modern societies. we're scum. dirty scum.

Re:the lengths people will go to...

[Anonymous Coward]

Yeah, society was much better back in the days when bakers used to adulterate their flour with ash or white lead - or when "gin" was sold that was really dilute sulphuric acid - or when tobacco firms covered up research proving they were selling cancer sticks - or when meat wholesalers decided to bulk up chicken by injecting it with water, beef protein and pork protein. Yeah, we're really awful now that we've progressed to optimising graphics cards to run specific benchmarks.

Re:the lengths people will go to...

[avalys]

Hopes and dreams? We're talking about video card benchmarks!

Re:the lengths people will go to...

[Machine9]

it's obvious isn't it? I DREAM of having great FPS and in the end I can only HOPE that the store will give me a refund...

Re:the lengths people will go to...

[Billly Gates]

He is complaining about the lack of ethics in society as a whole.

I agree.

This current time is looking alot like the late 19th and early 20th centuries. This is when the divide between rich and poor was about as extreme as today. 99% of the worlds money is owned by 1% of the people! Ronald Reagan started this craziness and after a taste of greed companies and the wealthy can not have enough. Before he took office the rich were taxed close to 90%! Today my taxes have gone up because I had to take delivery driving because the .com bubble collapsed. But gets who gets the tax breaks at expense of mine? Supprise Larry Elison, Bill Gates, Dick Cheney extra. They already have way fucking too much.

Its also when companies began sweatshops and looking for cheap labor. Child labor and no regulation controlls were the norm.

But the most obvious thing is monopolies are rampant today as they were back then. Back then we had oil and railroad monopolies. Today we have conglomorate media companies controlling what we see and hear. The fcc is in the corrupt pockets of these corporations. We also have Microsoft ruling with an iron fist and killing innovation. THe only competition is free software which brings the value of software and our IT wages down.

I blame it for part of the reason I had to leave IT. I use FreeBSD myself but if MS wasn't so oppresive and wiped out so much competition its likely that I could buy a cheap unix today for the same cost of windows.

But back to what the other guy is saying is true and sad. No ethics but just bottum line and extremely short time profits for Wall Street and and the lack of distribution of wealth is the norm!

Cheaters never win!

[Anonymous Coward]

Just ask Microsoft.

Oh...wait...

Never mind...

Re:Cheaters never win!

[FroMan]

I believe what you were thinking is:

Cheaters never prosper, but they win a lot more often.

Re:So What? Who Cares?

[jthj]

Nope not high enough :-P

High or low level strategy?

[MattGrounds]

What I'm interested in is:

a) Is this indicative of a high level strategy by NVidia's management, who's marketing department is pressuring them to have higher 3DMark2003 scores than ATI?

OR

b) Has some low level device driver programmer (intern?) looked to get some easy brownie points by "optimising" the drivers for 3dMark2003 in a slightly clunky way?

Either is quite interesting :) I've been a victim/perpetrator of both in the past.

Re:High or low level strategy?

[somberlain]

Well... All I know is, that a lot of companies do this... This is also the case for CD-R/RW-writers. When using tools like Nero CD-Speed etc, this problem is always solved by 'cheating' on the program. Been there for a while in this sector of hardware testing...

Re:High or low level strategy?

[BigFootApe]

a) Is this indicative of a high level strategy by NVidia's management, who's marketing department is pressuring them to have higher 3DMark2003 scores than ATI?

Of course it is. Fudging the drivers for a synthetic benchmark are a time honored way to make crappy hardware look good.

Re:High or low level strategy?

[Anonymous Coward]

I have been a victim of such things in the past, but I have never been a perpetrator, in spite of pressure to do so. I believe my exact words were "I don't need this job that bad! Fire me!" I didn't get fired and I didn't fudge the results.

Damnit, we all, individually, need to take this stance. Then maybe, collectively, the problem will go away!

Change the executable name...

[Anonymous Coward]

to Quack3.exe and see what happens...

Re:Change the executable name...

[Anonymous Coward]

... and don't forget to duck!

Make NVIDIA drivers Open Source!

[Lieutenant_Dan]

If we could have the Open Source developer community review and improve these drivers we would not encounter any problems with them. The experience and integrity of the Open Source developer community would be vital for the consumer to take Nvidia cards seriously in the market.

Benchmarks would reflect the actual performance of the card instead of skewing the results in order to garner favorable reviews.

Only when we allow Nvidia to see the benefits of Open Source can we free the graphic benchmark software from the clutches of Matrox.

Re:Make NVIDIA drivers Open Source!

[iainl]

Lovely idea in theory.

In practice, nVidia have made it painfully aware on numerous occasions that they CANNOT do this. Its not just them being nasty closed-source meanies. The driver binaries contain licensed tech from numerous third parties that their license doesn't let them reveal the source to.

Re:Make NVIDIA drivers Open Source!

[confused one]

Never happen.

Any technical advantage their hardware gives them would be lost; because, they'd have to reveal the specs for all to see.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

The driver binaries contain licensed tech from numerous third parties that their license doesn't let them reveal the source to.

That's right, so Nvidia should instead release specs (not under an NDA) on their chips, just like almost every other IC manufacturer in the world.

Re:Make NVIDIA drivers Open Source!

[alienw]

Go make some drivers and STFU. Then you might realize that writing drivers takes effort. Also, read the article. ATI was "examining" the drivers when they found the cheat. What, does nVidia want to make it easier for them to reverse-engineer by providing full source / specs? Think a little before whining.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

let's say I buy that old crap... what's the excuse for not releasing specs or low-level programming documentation?

Re:Make NVIDIA drivers Open Source!

[Arker]

Hello?

We don't want their stinking source code. We want the specs. Even if their binaries are encumbered, the specs aren't.

Re:Make NVIDIA drivers Open Source!

[3dr]

Ha ha, good bit of satire!

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

Yeah sure!! They "CANNOT" do this because of these nasty patented third-party cheatshhhh optimizations! ...oh, wait... now i know : maybe there's some SCO code in nVidia drivers??!;o)

Re:Make NVIDIA drivers Open Source!

[Richard_at_work]

Yes, they are. There is as much in the cards firmware that is licensed as there is in the drivers.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

Odd. SGI were the ones touted as having the undisclosed IP in the early days. When asked an SGI spokesman say "nope, not us. anything that they have of our IP that we know about can be disclosed if they want to".

How can the driver information have the IP shown in it, anyway? Unless a lot of the card functionality is hidden/exposed in software.

Otherwise it's as closed IP as the DirectX API Or OpenGL.

I reckon that the capability of the cards in the GF4 and earlier were tuned down so that the "professional" versions could work better.

Either that, or they have somebody else's IP without their knowledge in there.

Re:Make NVIDIA drivers Open Source!

[ImpTech]

Yeah so they say... they also say that they don't use their drivers to cheat at benchmarks.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

I cannot agree with you more, but unfortunatly the gfx card market is a money maker, and as we all know ; offering the code for free will allow the competitor free rule on the inner workings. Normaly this would not pose a proble,BUT most graphics intense programs are on Windows-not Linux.

Re:Make NVIDIA drivers Open Source!

[osu-neko]

Go make some drivers and STFU. Then you might realize that writing drivers takes effort.

Are you replying to the correct article? Where did the person you are replying to say or imply writing drivers was effortless?

Also, read the article. ATI was "examining" the drivers when they found the cheat. What, does nVidia want to make it easier for them to reverse-engineer by providing full source / specs?

That would not be the goal nVidia would be shooting for, rather an unfortunate side-effect, but yes, absolutely, this is something nVidia should want done. Did you read the article? Did you notice the blatantly obvious fact that hiding their source code did not in any way prevent ATI from reverse-engineering their drivers? So what benefit do they get from hiding the source? (Actually, one fairly obvious one -- it allows them to easily hide the fact they're cheating on benchmarks. But assuming everything was on the up and up, what would the benefit be?)

Think a little before whining.

A little more politeness would go over better, especially when it's your own post doesn't seem to have been fully thought through...

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

There is as much in the cards firmware that is licensed as there is in the drivers.

*sigh*

Once again, we are talking about the specs for the chips, not firmware and not software. We want to know the different registers address, size and functionality. Anything more is just a bonus.

There is nothing patented or copyrighted about that (and no, the UDA patent does not cover this).

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

Knowing the specs of the chip won't prevent cheating on benchmarks with drivers, which is what the whole point of this part of the thread is about.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

A better alternative:
don't buy and don't use nVidia products.

Re:Make NVIDIA drivers Open Source!

[damiam]

ATI manages to release specs, so there has to be some way around it.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

Then GPL it. That way nobody can use it without giving back.

Re:Make NVIDIA drivers Open Source!

[Arker]

Au contraire.

Open source drivers will prevent cheating (you can't hide cheats in there when people can see the source,) and open specs will mean that good open source drivers can be written, which can then be used for all applications, including benchmarking.

Re:Make NVIDIA drivers Open Source!

[R0]

The Benchmarks would need to be open too. As would the operating system :)

Re:Make NVIDIA drivers Open Source!

[alienw]

Reverse-engineering drivers is one of the most difficult things you can possibly do. I can assure you that ATI cannot heavily reverse-engineer the drivers -- that's too hard.

First, even a 20 kbyte DOS program is a bitch to disassemble and understand. A 5 MByte driver is pretty much impossible to understand at all from the raw disassembly. Second, disassembly is usually prohibited by the driver EULA, so you have to tread carefully. Third, reverse-engineering is prohibitively expensive due to the time and expertise required and ATI likely does not do a whole lot of it.

In short, not having the source pretty much stops most reverse-engineering attempts, for both technical and legal reasons. Hell, with the source you don't have to do any reverse-engineering. You just read it and re-use the good code in your own open-source driver, thereby saving money and undercutting the competitor.

I think ATI would get a huge competitive advantage if they could just download the competitors' drivers and modify them to work with their own cards. I'm sure that pieces like the GL implementation are quite reusable. So, where is the competitive advantage for nVidia here? The only people who benefit are the consumers interested in using the source -- clearly not a large group, given that even most Linux users don't particularly care if they have the source code.

I sure as heck wouldn't mind having open-source graphics drivers, but to expect that is downright unreasonable. The field is too competitive for that to happen.

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

Did they licensed cheating tech from ATI?

Re:Make NVIDIA drivers Open Source!

[Anonymous Coward]

ATI manages to release specs

Really, so where can I download the specs? What, I have to sign an NDA?

The specs have to be released to the general public, or they might as well not have been released at all. Specs under NDAs are worthless (as is evident by the quality of the opensource ATI drivers, i.e. buggy and lacks alot of features).

If you want to see how you properly release hardware specs, then got to Texas Instruments or National Semiconductors website. Why graphics chips hardware companies are incompetent in this respect, is beyond me...

Renaming

[jabbadabbadoo]

I tried various tricks too. Oddly, renaming it to Outlook.exe made it crash.

Re:Renaming

[crawling_chaos]

Oddly, renaming it to Outlook.exe made it crash

It also sent an e-mail with "I Love You" in the subject to all of your contacts. I wonder if the reason for delays on the FX card was that they'd renamed the burn-in application DukeNukemForeever.exe?

Thank you, thank you. I'll be here all week.

Re:Renaming

[Carbonite]

I wonder if the reason for delays on the FX card was that they'd renamed the burn-in application DukeNukemForeever.exe?

Damn, NVidia's going to be pissed. They delayed the launch of the FX cards and misspelled the file name.

Re:Renaming

[llamalicious]

It's KarmaBurnFriday(tm)!

SWEET! Where's the party??????

Oh, that's right... in my pants!

Re:Renaming

[Inda]

Well fucking fuck my fucking karma today as well then.

BTW you all suck.

Re:Renaming

[grub]

In your pants? Yahoooo! 6 days of getting karma to "Excellent" (nee: "50") just to FUCK IT ALL UP IN ONE DAY so For The Record:

Linux sucks
Windows sucks
RMS sucks
Linus sucks
Microsoft really sucks
Apple sucks
*BSD sucks (and is also dying)
slashdot sucks
everyone on this fetid, shithole of a planet sucks.

Thank you.

Re:Renaming

[IPFreely]

Oddly, renaming it to Outlook.exe made it crash.

Actually, that is very interesting. Any bets you've stunbled onto an entirely different cheat?

Let's say Microsoft wanted Outlook to have some special capabilities in the operating system. So the OS recognizes it and gives it special treatment. Another app comes along with the same name and triggers the "special treatment" but can't handle it. Ka-Boom.

This also brings up the possibility of really screwing with these drivers. Go get another game program (QUAKE.EXE or whatever) and rename it to the name of the benchmark. What does the driver do to it?

Re:Renaming

[Anonymous Coward]

So if I rename armyops.exe to 3Dmark.exe, will I get better graphics performance in America's Army?

Re:Renaming

[edwdig]

Microsoft does give special treatment to special executables. I never tried any testing to see what exactly triggered it, but do the following:

Install Windows 98
Delete the Internet Explorer icon off the desktop
Install Netscape Communicator 4.x

Look at your desktop - the Internet Explorer icon will be back.

Re:Renaming

[BrynM]

say Microsoft wanted Outlook to have some special capabilities in the operating system

Then you would be saying the truth. Outlook, and the rest of Office, register themselves extensively into the OS. To see what I mean, grab scanreg.exe (if my site goes down, look on Google) and run the command "scanreg outlook". It will show literally hundreds of entries in the registry for both Outlook and Outlook Express (on my machine 1157 occurances). For comparison if you have Norton Anti-Virus (which should register itself extensively into the OS), try the command "scanreg norton" and note the 150 or so results (163 on my machine). I'm betting more than Outlook will eventually crash with constant use of the machine and a re-named outlook.exe.

Re:Renaming

[damiam]

If you rename the fairy.exe file from Nvidia's Dawn demo to quake3.exe or 3dMark03.exe, Dawn'll be naked.

Re:Renaming

[Anonymous Coward]

you mispelled the word mispell retard.

Re:Renaming

[eatdave13]

*starts moshing*

KARMA KEG STAND!!!

FUCK KARMA!!! I HATE FUCKING KARMA!!!

Fuck Linux.

Embracing MICROS~1 extends my wang.

I hope...

[Anonymous Coward]

they didn't rename it to quake3.

I will take a wait and see attitude.

[Prince_Ali]

You can't trust results for older games, and you can't trust benchmarks evidenlty. I think the best thing is just to wait for the new generation of games which will surely clear things up.
It doesn't make sense to buy a card to run Doom 3 when the game isn't out. Here is a clue, when Doom 3 does come out I will be able to buy something as powerful as the FX 5900 for $150.
I'm going to go into an offtopic rant now. It is sad that we have huge displays and crazy-go-nuts graphics processors on computers, but consoles will probably always beat PCs for game size. Game makers are too scared to release a DVD only game, so our games are limited to 700MB by disk, and don't even get me started on controllers.

Re:I will take a wait and see attitude.

[ShieldW0lf]

Obviously you never played Planescape: Torment (5 discs)

Re:I will take a wait and see attitude.

[Anonymous Coward]

Console controllers are the worst game-control devices ever invented.

PC games can be, and often are, far larger than console games.

What on Earth were you talking about?

Re:I will take a wait and see attitude.

[PhxBlue]

Whatever happened to Origin? I can remember three distinct occasions where I upgraded my PC just so I would be able to play an Origin game (Wing Commander III, IV, and WC Prophecy). Point is, if you release a good enough game, people will upgrade their PCs to get it.

Maybe we'll see Doom3 on DVD? It's a thought.

Re:I will take a wait and see attitude.

[Prince_Ali]

Look at Enter the Matrix. On an XBox or PS2 I assume you get the game on one 4 GB disk. On GC you get it on two 1.8 GB disk. On the PC you get it on 700 MB disks. Am I the only one who sees a problem here? To get decent sized games they have to release it on 3-7 disks (think Neverwinter Nights, Baldur's Gate the list goes on). If you don't do a mind boggling full install you have to switch disks all the time. I just want a bold game company to release its games on DVD so we don't have this disk spanning issue. The things are what $30?

Also when I said controller I meant gamepad. I know that mouse/keyboard is best for FPS and strategy and many types of very good games (not to mention flight-sticks, racing wheels for other types of games), but PC gamepads stink. I wish MS would just sell X-Box-S controllers with a standard USB plug and drivers (I am aware you can do it yourself, btw).

Re:I will take a wait and see attitude.

[Anonymous Coward]

A "mind-boggling" full install? What, we're talking all of 4GB here? I'm running it on a PC --- so I've got a decent-sized hard drive!

PC gamepads stink, yes, but this is a specific case of a more general theorem: gamepads stink.

Re:I will take a wait and see attitude.

[Prince_Ali]

That is exactly what I'm hoping for.

Re:I will take a wait and see attitude.

[untaken_name]

Console controllers are the worst game-control devices ever invented.

For first-person shooters, maybe. (and some RTS games, of which very few make it to consoles)
The DualShock2 is the best controller out there for just about everything else. It's comfortable, it has intelligently placed buttons, it has enough buttons, but not *too* many, and for me, the most important part is that I forget I'm holding a controller when I'm using it. It helps me 'just play' and not be distracted. I agree that there are a few game types it won't work for, but then again it wasn't designed for those kinds of games. (for example, it's not that great for gun games.) PSjoys and other console-to-pc controller adapters would not have market share if people *only* wanted to use mice/keyboards or the sidewinder gamepad, which is inferior to the DualShock2 in every way.
Obviously, this is just my opinion, but I believe it to be a valid counter to your opinion.
This is a bit OT, but all I could think of when I read your comment was the comic book guy from the Simpsons pointing at me and saying 'Console controller? Worst device EV-er!'

Re:I will take a wait and see attitude.

[Coyote67]

Scared to release a dvd only game? How about smart not to. Believe it or not, not everyone has a dvd drive. And what the hell do you care what a game comes on if its the same game? Is your life so important that you can't swap the discs 4 times so you can play enter the matrix?
I'm not trolling or anything I'm just tired of people complaining about the game industry not releasing on dvds, its always the same people who also complain about them pushing newer hardware.

Re:I will take a wait and see attitude.

[TwistedGreen]

oh yes, and we all know how beautiful those huge high-res textures look on a TV.

Re:I will take a wait and see attitude.

[Prince_Ali]

It had a real effect on Enter the Matrix. The in game movies actually looked better on console because they didn't have to compress so aggressively. I personally don't like keeping track of 3-7 game disks for every game especially when they all come in those cheap paper sleaves.

Just to clarify I love PC gaming, but I just believe it has a couple weaknesses. This particular weakness could be addressed with a $30 dvd drive. When most gamers spend $100 for a soundcard and even more for a videocard it makes sense.

Re:I will take a wait and see attitude.

[thirdbanE]

I think that keeping the games on CDs is a good thing, it keeps the developers from plunking down 10 GB games on users' home hard drives. DVDs may be big, but they're still not faster then CDs are.

Re:I will take a wait and see attitude.

[Prince_Ali]

Are you sure that they aren't faster? I really don't have any numbers for new DVD drives, but my dvd drive is only 2nd or 3rd generation and it maxes out at around 5.5 megabytes per second (around 40X in terms of CD speed).

Re:I will take a wait and see attitude.

[johndiii]

Part of this is distribution media keeping pace with fixed storage media. Back in 1985, fixed disks were 5-10MB, and distribution media (5.25" floppies) were 360K. You could fill up your 10 MB disk with 28 floppies.

Today, your 120GB disk would require over 170 CDs (at 700MB each) to fill up (leaving compression aside, this is just a back-of-the-envelope estimate). By contrast, a 4GB DVD brings the ratio to 30 disks. Roughly comparable to the numbers 18 years ago. Of course, we are moving on to larger disks; 250GB is now the top end.

Due to the antics of copyright holders and media/drive maker "consortiums", the rate of expansion of distribution media has fallen behind that of fixed storage media. The DVD has still not become the accepted distribution format, despite the fact that DVD drives are nearly as cheap as CD drives, and DVD writers are becoming affordable.

It's time for the computer and software manufacturers to get their act together and move to the next logical step for distribution.

Re:I will take a wait and see attitude.

[Lightwarrior]

> Whatever happened to Origin?
http://www.origin.ea.com/
Technically, they're still around. However, after being purchased by EA (IIRC), they lost (EA fired) the majority of their creative and technical talent. The result? Origin exists only in the capacity of endless UO expansions.

> Point is, if you release a good enough game, people will upgrade their PCs to get it.
This is no different on the PC than it is for anything else. Houses, cars, console systems, PC, guns, ceiling fans... when one doesn't cut it anymore (defined by what action you wish the object to complete), you repair/upgrade/replace it.

PCs in particular have an interesting co-dependant relationship between software and hardware. More demanding software products yield better hardware (ie, benchmarks), while more powerful hardware allows "better" (faster, prettier) software.

> Maybe we'll see Doom3 on DVD? It's a thought.
Why? What's the advantage of going DVD over CD for the majority of games released? Unless we're talking Baldur's Gate-style 6 CD sets, what's the incentive to switch?

There are CD cases that contain two CDs, for the ones that still come in jewel cases. Neverwinter Nights, which came on three CDs, had the disks in a CD envelope instead of a case.

And since hard drives are so large, the only reason most games need a CD in the drive is to make sure you've still got 'em.

Think of it this way: what incentives do publishers have to release DVDs instead of CDs?

-lw

Re:I will take a wait and see attitude.

[Talez]

The in game movies actually looked better on console because they didn't have to compress so aggressively.

Game movies always look better on a TV screen. The nice low resolution display gives it a little bit of poor man's interpolation. Meanwhile, the crisp hi-res display of the PC means you can see every artifact in its ugly glory.

Re:I will take a wait and see attitude.

[afidel]

Fixes disks are dirt cheap for huge capacities and are orders of magnitude faster than optical, especially for seek time. Even if a game came on DVD I would still do the full install if for nothing other than faster area transitions and the lower noise level of not having my optical drive spinning all the time (my cdrw in particular is the loadest thing in my PC by about 20X (it's quite a few Db louder than the case fans).

Re:I will take a wait and see attitude.

[damiam]

Here is a clue, when Doom 3 does come out I will be able to buy something as powerful as the FX 5900 for $150.

The Radeon 9700 PRO still costs ~$285, almost a year after its release. You might have to wait a while to see the FX 5900 at $150.

Re:I will take a wait and see attitude.

[Anonymous Coward]

Or Metal Gear Solid 2: 7.5GB full install.

Re:I will take a wait and see attitude.

[aardwolf204]

Game makers are too scared to release a DVD only game, so our games are limited to 700MB by disk, and don't even get me started on controllers.

Wait and see. Remember how long it took AOL to start spamming your mailbox (physical) with AOL CD's instead of AOL Floppies? Sure most computers now come with DVD drives, but if even 10% of the games audience only has CD drives then the company will loose money. I dont think they're ready to make that plunge until the water is very safe. For now 2 or 3 650MB CD's do the trick. So what if it costs a few more pennies on production costs when its a $50 game.

And even when they do start releasing games on DVD, and trust me, they will, they will probably have a way for you to order it on 5 CD's.

I remember years ago, right after AOL started sending out CDs, I would call up AOL and ask for them to send me a copy on floppy. A few weeks later I would have 6 shiney new 1.44meggers in the mail ready to be formatted and have Doom .wad backups put on em. But then again, that was prolly before your time. =)

ATI Drivers SUCK!?? Oh wait..

[rkz]

I hope this puts an end to the urban myth that ATI drivers suck!
I'm finding that the 4x.xx drivers don't work that well with my GForce 440MX the TV-Out is screwy, the best drives for me are 30.82 I wish they would fix that instead of optimizing for pointless benchmarks! I'd rather have a driver that works than one that gives me a bigger 3dMark score!

Re:ATI Drivers SUCK!?? Oh wait..

[Anonymous Coward]

Not sure on that as my Radon VE drivers seem to crash XP all the time.......er,obiously I'm no games nut........but I like to play solitaire and that always gives me a BSOD :((((((((

Re:ATI Drivers SUCK!?? Oh wait..

[ananke]

your logic is flawed. just because nvidia's drivers may suck for you, does not result in ati's drivers sucking less.

Re:ATI Drivers SUCK!?? Oh wait..

[afidel]

sorry to hear it, I had a spat where the dets weren't allowing my pc to resume from sleep correctly. But the latest WHQL drivers fix that again AND gave me an ~20% performance gain in NWN to boot. Try the 43.51's and see if they don't work well for you.

Re:ATI Drivers SUCK!?? Oh wait..

[rkz]

I'll give it a try thanks

Who cares about benchmark software?

[i.r.id10t]

Really? I mean, I've been playing 3d FPS games since the original Quake/QuakeWorld, and the only thing that has mattered since then is what kind of score you can get on a timedemo. I don't care if a card can get 200000 frames/sec in glxgears or the 3dmark tests - I care about how it works in the latest/greatest 3d FPS. In Quake/QuakeWorld, you *needed* to get at least 40 frames/sec. In Quake2, 60 was an ideal minimum. With Quake3, it changed to 125f/s because there were/are some trick jumps/moves you can only do with a minimum of 125f/s framerate. And of course, when playing online, your connection had to be able to get enough data to feed the card as well.

So throw the benchmark software out, fire up Q3 or whatever, and let us know how the card really performs.

Re:Who cares about benchmark software?

[Trolling4Columbine]

"With Quake3, it changed to 125f/s because there were/are some trick jumps/moves you can only do with a minimum of 125f/s framerate."

I hate to tell you, but framerate has NOTHING to do with what the game's engine will let you do.

And if I recall from other discussions here, the human eye/brain cannot process much more than 30-35 FPS.

Re:Who cares about benchmark software?

[Toraz Chryx]

bzzt, quake 3's physics are tied to the framerate, you ca jump slightly higher with a higher framerate. ergo, with 125fps+ you can get to locations via jumppads that are inaccessible otherwise.

and the fps thing.. well, you're just outright wrong on that, I can easily see a difference between 30fps and 120fps :)

Re:Who cares about benchmark software?

[Tipsy McStagger]

the q3 physics when jumping works best with fps's where the rounding errors in calculating the path through discrete points are maximised.

43 76 & 125 all produce similar results.

http://ucguides.savagehelp.com/Quake3/FAQFPSJump s. html

Re:Who cares about benchmark software?

[10Ghz]

it matters because companies use 3DMark in determining what vid-cards to put in their systems. It matter because there are no other benchmarks besides 3DMark that let you benchmark features that are not yet found on games. That's the point of 3DMark: To test features that are not yet aailable in games. Games are lagging behind when it comes to implementing features. Hell, Doom 3 is targeted at GeForce's feature-set (you know, the ORIGINAL GeForce. After that we have had GeForce 2, GeForce 3, GeForce 4 and GeForce FX)!

Re:Who cares about benchmark software?

[Trolling4Columbine]

"quake 3's physics are tied to the framerate"

So Joe Schmoe with the Geforce2MX getting 50FPS has a built-in (read: deliberately programmed into the engine) disadvantage relative to the guy with the new Radeon 9700? Somehow I don't think Id would intentionally sabotage players using slower machines.

"I can easily see a difference between 30fps and 120fps"

But your eyes/brain can't distinguish each individual frame at that high a framerate... that's why insanely fast timedemos appear to be a blur.

Re:Who cares about benchmark software?

[Anonymous Coward]

So Joe Schmoe with the Geforce2MX getting 50FPS has a built-in (read: deliberately programmed into the engine) disadvantage relative to the guy with the new Radeon 9700?

No, he has a built-in ACCIDENTALLY programmed into the engine disadvantage relative to the guy with the new Radeon 9700? Unless he drops his framerate to 43. It's due to rounding errors being more significant at certain rates.

Re:Who cares about benchmark software?

[Spellbinder]

does this make any sense???
when finally a game comes out that uses your graphic card features the card will be way to slow
i mean common, would you play doom III on a geforce (I)
a long time DIII was aimed at gf3 but i would like to buy a card faster then a radeon 9700 Pro to play it
i say what matters are the games that come out in a year after the card is bought
but 3DMark aims at features which will be in games in 2 or 3 years

Re:Who cares about benchmark software?

[entrager]

Actually the limit is closer to the 70-80 Hz range. Try changing your (CRT) monitor's refresh rate to 60Hz, most people can easily see it flicker at that rate (especially large white areas). However, at 72Hz most people can no longer see the flicker.

I KNOW 125fps is way above the human limit. I tend to disregard any comments people make about being able to tell the difference between frame rates higher than 60. While we are technically able to tell the difference between 60 and 70, it's too damn hard to during a game in which things are changing too quickly.

Re:Who cares about benchmark software?

[Bartmoss]

...except that ATI has been known to "optimize" for Q3 as well. A benchmark is a benchmark, by any other name it can be cheated just as well.

Re:Who cares about benchmark software?

[afidel]

Well since they give average framerate a number higher than 60fps CAN be usefull. The reason is is that at a higher average framerate the worst case is likely to be better than a card that can barely maintain the 60+fps needed for completely flicker free gameplay. If more games and reviewers gave minimum framerates then we wouldn't have to rely so much on the rediculously high average FPS to know we are going to get decent worst case fps. As an example I get around 42fps in NWN at my lcd's native resolution and with most effects on and max details, but when a bunch of spells are being cast and other things are going on onscreen (translucent clipped textures mostly) I can get down well below 20fps which is noticably choppy.

Re:Who cares about benchmark software?

[El_Ge_Ex]

Actually, it's not quite that bad. Doom 3 is ahead of the rest and aiming for GeForceFX/Radeon9700. The minimum will be GeForce2/Radeon8000. Anything earlier will just not have the features and software rendering will be too slow.

Don't forget, there will be an Xbox (Basically GeForce3) Version of Doom3, yet with the PC version it still isn't a good reason to buy one!

Dammit, I would have had an Xbox months ago if they just made a decent number of exclusive games fot it!

If I were Microsoft, I would be putting ALL my Xbox resources into getting some damn games out!

-Brian

---
No, it's not off-topic. It's Anti-Microsoft and this is Slashdot.

Re:Who cares about benchmark software?

[Kashif Shaikh]

So you mean to tell me, I just have to set my refresh rate to 76hz /w vertical sync and I can jump higher then the little 12-year brat?

I guess it helps to look down on the floor, so that your frame rate doesn't drop below 76fps while you jump...

Weeeeee.....

Re:Who cares about benchmark software?

[Toraz Chryx]

Did I say I could distingush each frame?..

no, and that's the entire point, it appears smooth in motion and not a bunch of one-after-the-other seperate images.

as for Q3, no, it wasn't intentional, Q3 wouldn't even RUN at 125FPS on the hardware available when it was developed.

Re:Who cares about benchmark software?

[Anonymous Coward]

Maybe so, but in most cases there's no way you could tell the difference between 60 and 120fps, because your monitor doesn't update the screen that fast.

Tell me about it...

[Anonymous Rockstar]

Just yesterday I installed my 8x geforce 4200 and I installed the new detonator fx drivers and My antialiasing wouldn't work, but it did before I loaded it. Nvidia drivers are usually know for putting out a great driver that gives a few more frames a second, but I'm not giving up my antialiasing.

So? What others are optimised?

[Matey-O]

If renaming the file changes the optimizations, why doesn't anyone (read: not me) take a gander at the detonator drivers and figure out what OTHER games it's tuned to?

Re:So? What others are optimised?

[Anonymous Coward]

So, you're saying that if Nvidia happened to have made optimizations for specific games - that it would be bad? Perhaps if the optimizations meant image degredation but, otherwise I can only see it as a good thing.

Re:So? What others are optimised?

[mosschops]

why doesn't anyone take a gander at the detonator drivers and figure out what OTHER games it's tuned to?

I've had a quick look at the ANSI and Unicode strings in NV4.SYS and NV4_MINI.SYS (the main NT/2K/XP drivers), as well as NV* files in the Windows\System32 directory. Despite lots of other strings being found, there's no mention of 3dmark or quake. I guess they must be making a special effort to conceal them - suspicious in itself.

If they are only making optimisations for speed, without loss of quality, would they have any reason to hide them?

As soon as someone does dig the information from the binaries, the applications could no doubt be checked for further cheating.

Shades of 3DFX in nVidia

[saden1]

If nVidia continues to have these bulky video cards which take two PCI slots and make noise like a whale they just might go by the way side just like 3DFX.

nVidia is walking a tight rope and for the first time in six years I'm actually going to consider buying an ATI. Come September 30th there is 90% chance that I'll have an ATI card on my machine.

Re:Shades of 3DFX in nVidia

[gordyf]

FWIW, I've had a Radeon 9700 Pro since December '02 and I love it. Used to be an nVidia guy, but I haven't looked back since. I was a bit worried about ATI's drivers, but the Catalyst series have been great.

Re:Shades of 3DFX in nVidia

[saden1]

ATI drivers have always been the deciding factor and now that people are actually saying good things about their drivers I feel a lot more comfortable buying an ATI card.

Re:Shades of 3DFX in nVidia

[ejaw5]

It's not like you're going to use that PCI slot next to the AGP anyhow. A lot of video cards come with fans and you wouldn't think of blocking airflow by sticking a PCI card up next to it.

Re:Shades of 3DFX in nVidia

[Spellbinder]

you are talkin about linux drivers, aren't you

Re:Shades of 3DFX in nVidia

[untaken_name]

FWIW, I've had a Radeon 9700 Pro since December '02 and I love it. Used to be an nVidia guy, but I haven't looked back since. I was a bit worried about ATI's drivers, but the Catalyst series have been great.

I switched to a Radeon AIW 9700 Pro about 2 months ago, and I am extremely happy with it. I am a *very* loyal video card buyer, and it was difficult for me to buy an ATI card. I stuck with Diamond as long as I could, but when I could no longer get a Diamond, I switched to nvidia (as my last Viper had an nvidia chipset anyhow) and stuck with them through the Riva, TNT, TNT2, Geforce, geforce2, (couldn't afford a geforce4, but I woulda bought one if I could have :) and I was going to be buying a GeforceFX, but luckily for me my geforce2 died before the FX series came out. Faced with the choice between no PC or an ATI card, (as I didn't want to mess with a temporary cheap card until the FX came out, too much trouble) I chose the ATI card, and I was just blown away. It's been rock solid, and unless nvidia really gets their act together, I won't be buying another card from them. I am no fanboy, I just prefer to work with companies I like and trust (to have a decent product). It seems nvidia is intent on knocking themselves off that list.

Re:Shades of 3DFX in nVidia

[saden1]

They barely had what you would call Linux drivers. I was actually talking about the Windows driver which where abysmal when it came to playing games. The drivers where poorly written, crashed often, and had graphics artifacts problems.

Re:Shades of 3DFX in nVidia

[FroMan]

Hmmm, I run a ATI Radeon 9000 Pro card in a box with an AMD Athlon 1700+ and it stays around 50-55c. Over all the box is pretty quiet.

While folks have claimed terrible problems with ATI drives I can't say I have too much to complain about. In windows I do fine with WCIII and a number of other 3d games. In linux enemy territory test had a lot of graphics problems, but other games didn't seem to fare so bad. The ET stuff though I think is more the game than the card/drivers though. The game in both win and linux are less stable than san andreas fault.

Re:Shades of 3DFX in nVidia

[Anonymous Coward]

Also, I've seen a few motherboards that don't even have a PCI slot next to the AGP. Mine has one of those worthless CMR or CNN or whatever slots in between the AGP and PCI slots. Might be something to look for in a new mobo if you're building a new system using all 5-6 PCI slots and can't live without that high-end game system/jet engine combo. ;)

Speaking of C** slots, does anybody actually use them? Practically every motherboard I've seen has one, but I've never seen anything to put in one.

Re:Shades of 3DFX in nVidia

[illumin8]

100% agreed.

I just bought a new Dell and for only an extra $110 on the system price I got a Radeon 9800 128MB (not the Pro, but it's the same card, just clocked 50mhz. slower).

This card rocks! I was looking at the benchmarks on the site linked to in this article and my mid-budget $110 Radeon 9800 is outperforming the GeForce FX 5800 by about 200 in the 3dmark scores...

I would highly recommend the 9800, not the 9800 pro, which is just too expensive.

Nvidia released their next generation of cards a day late and a dollar short, and now they are having to cheat just to make it look like they can keep up with ATI. This ATI card only has a small fan on it and makes practically 0 noise. It is one great card and I will probably not buy an Nvidia again. I had a GeForce 4ti4200 in my last system and it was a good card, but the capacitors started to leak and the card went bad after about 6 months... :-(

Re:Shades of 3DFX in nVidia

[CoreDump01]

The game in both win and linux are less stable than san andreas fault

I'm running ET/Linux on a P4 with a G4Ti4800 and it's rock solid.

Re:Shades of 3DFX in nVidia

[saden1]

So in effect an nVidia card would need 3 slots (1 AGP and 2 PCI) because you wouldn't think of blocking the air flow. With a card that only requires one slot you would have more freedom because you can have it take up two slots (1 AGP and 1 PCI) and still have more air-flow space than an nVidia card that reqires two slots to begin with.

Your logic is flawed.

What can we do to help?

[Anonymous Coward]

I really feel for you. Is there a way that the /. community can assist you with your problems? Can you list the specs of your system?

Re:What can we do to help?

[Anonymous Rockstar]

I just rolled the drivers back. It's not a big deal. I just expect more from Nvidia. I guess I stuck a thorn in the side of a NVfanboy.

Re:What can we do to help?

[Anonymous Coward]

Actually my card is a Trident VESA card. I don't own anything from the big four: Nvidia, ATI, Matrox, and Jim's Video.

Glad to hear it works again.

Take care ...

Re:What can we do to help?

[Anonymous Rockstar]

Hey! I used to own a Jimforce 3 Ti200. I hope your not knockin the Jimforce series. ;)

Re:So What? Who Cares?

[confused one]

Uh, no. The mass market will buy the box with the prettier logo, The box with the lower price, or the box that magazine xyz (that they read) says is best.

Sad, but true.

Most people will shop around, to make sure the features they're looking for are simply there and work. Beyond that, they don't do the research to understand which version is better unless they're forced to.

Re:So What? Who Cares?

[Anonymous Coward]

>> Aren't we smart enough not to be pulled in my marketing hyperzor?

No.

Re:So What? Who Cares?

[Eric_Cartman_South_P]

Aren't we smart enough not to be pulled in my marketing hyperzor?

You are... I am... And a bunch of /. readers are too. But the millions of teenagers that grew up without technical skills but love games and subscribe to gaming magazines are looking at the benchmarks to decide what to ask for X-Mas and what to beg Mom to buy at CompUSA. I've seen it happen in person with a girlfriends younger brother. The difference between Quake II scores of 110 FPS to 112 FPS was a world of difference. Of course, the card with another 2 FPS was bought! (Not actual numbers, but I remember the difference was in fact 2 FPS!)

That said, what NVidia is doing is cheating, plain and simple. No laywer or press release can spin it otherwise. Well, they try, but the truth hurts in its simplicity. Change the .exe name and the cheats dissapear. And they are not "optimizations" because when the cheats are working they reduce the quality of the rendered image.

Lies, damn lies, and benchmarks

[EriDay]

Rather than an Nvidia problem, this is a benchmark problem. I don't know why people keep crying about this rather than fixing the benchmark.

Why isn't the benchmark a supervisor that renames the real benchmark to some random name, then runs it.

Seems to me the trick is to stay one step ahead of the marketers.

Re:Lies, damn lies, and benchmarks

[Anonymous Coward]

>Why isn't the benchmark a supervisor that renames the real benchmark to some random name, then runs it.

Yes, but then the nvidia driver could just check to see if the program that's being run has a random name!

Re:Lies, damn lies, and benchmarks

[ndnet]

No good. It could also check, offhand, process name, filesize, gfx calls, textures, checksums, etc. It's hard to disguise that well, and ultimately should they have to?

new name

[slide-rule]

Lemme guess... calling it 3DLark03 or 3DFark03? ;-)

Topic icon ?

[mirko]

I just noticed the Topic Icon was not ready for this story, so they put its name instead :-)

console controllers

[aliens]

Don't even get me started on console controllers, man do they suck, when will they just bundle a mouse and keyboard with these things.

Comment removed

[account_deleted]

Comment removed based on user account deletion

NVidia doesn't make the cards.

[Anonymous Coward]

Unlike ATI, NVidia sells chipsets to 3rd party manufacturers like ASUS, MSI, etc. to make the cards.

You can't blame NVidia for what kind of fan ASUS decided to use on their cards.

Re:NVidia doesn't make the cards.

[saden1]

Point is nVidia shouldn't have designed a chip that likes to be so hot that a big ass fan is required to keep it going. I'd like to go on record and say that nVidia's design is not natural and therefore inferior. One has to always strive for the elegant solutions otherwise we would all be using a computer the size of a refirerator.

Re:NVidia doesn't make the cards.

[IncohereD]

nVidia produces the reference designs, however, and this is where the big-ass fan came from.

Re:NVidia doesn't make the cards.

[Anonymous Coward]

The onus could just as easily be on the vendors to implement a more elegant cooling solution.

Or on the motherboard manufacturers to accomodate today's graphics cards.

Your logic is flawed.

Re:NVidia doesn't make the cards.

[Spellbinder]

WRONG!!!!
ATI is selling their chipsets too
but they have a ATI brand card too

Re:NVidia doesn't make the cards.

[Anonymous Coward]

Good thing that wasn't his point!

(fucktard)

Re:So What? Who Cares?

[s20451]

Forget features. I'm stuck with the vid cards that work under XFree86, in non-framebuffer mode. I still remember the day I tried to install Linux on a machine with an ATI Rage 128 card, only to have it say: sorry, driver not written yet.

For a good explanation...

[epicstruggle]

as to why synthetic benchmarks are useful please see the following:Beyond3D. This website is probably the best site for info on 3d hardware.

later,

Re:For a good explanation...

[Torgo's Pizza]

And as a counter, please read this article explaining why 3DMark is not a good benchmark because of the aggregate number it uses.

Re:For a good explanation...

[Anonymous Coward]

anyone who will rely on/praise Kyle Bennett from HardOcp for his "expert opinion" on anything hardware related, has lost even the slightest claim for objectivity. as such, your counter-viewpoint is sorely lacking.

perhaps you shouldn't dismiss beyond3d as easily. there's more knowledge and true 3D-industry experts assembled there than anywhere else on the web.

Baldurs gate...

[Anonymous Coward]

had 5, BG2 had 4, phantasmagoria had a bunch too, 6 maybe, BG1 was released on dvd the same year it was released on cd, and yet since then, how many computer games have been released on dvd?!? How many years ago was that?

By the time...

[Anonymous Coward]

It doesn't make sense to buy a card to run Doom 3 when the game isn't out. Here is a clue, when Doom 3 does come out I will be able to buy something as powerful as the FX 5900 for $150. ...Duke Nukem Forever comes out you can get one for $25.

Sweet Irony

[zeus_tfc]

Does anyone else see the irony in the fact that the "Graphics" topic doesn't have a picture?

Re:Sweet Irony

[_xeno_]

Yeah, I was wondering about that too. But then the next story seems to have no icons on the main page too.

It appears to be a glitch in the SlashCode.

Or, more likely, they're making changes to it again without bothering to alert any of the readers about them. I rather liked getting an explanation of the changes, even if it was the fourth SlashBack item out of five and no one ever read it. This new "change it as we go" thing is starting to get annoying.

(And it would be nice to get offical explanations of things like when some non-subscribers were treated to early story access... This "keep the users in the dark" thing that seems to be going on just makes it that more unlikely that I'll ever subscribe...)

Hmmm... I'll probably get modded down for this, but it's been bugging me for a while and I've got karma to burn. (Haha, with this sentence, I'm sure to get modded up! Whoops, didn't mean to think that out loud.)

Re:Sweet Irony

[Mordaximus]

Ironic, yes. But keep in mind that this is not cheating, it is simply an optimization to make the page render more quickly. Note however that if you renname the url to http://www.microsoft.com the optimization dissapears!

Re:Sweet Irony

[lordfoul]

"Does anyone else see the irony in the fact that the "Graphics" topic doesn't have a picture?"

Sure it does.. Oh you must Have NVidia Drivers installed.. Thay are only Optimized for Slashdot.com not Slashdot.org ..

Sorry...

Re:Sweet Irony

[happyhangone]

Maybe is my nvidia's card driver that is optimizing to speed up the refresh rate of the /. page...

quake3.exe

[Tom7]

I'll bet if you rename quake3.exe to something else, it also gets a bunch slower...

No Graphics Icon?

[crashnbur]

Does anyone else think it's a bit ironic that the link from the main page to this story, a story in the "Graphics" section, is missing its topic image?

Re:No Graphics Icon?

[Anonymous Coward]

Not ironic: irony is the difference between the literal and intended meaning. A more apt word would be 'significant.' But your point is well-taken. There's probably an inverse relationship between how often you get laid and how many fps your graphics card can do.

Re:No Graphics Icon?

[crashnbur]

Irony only refers to "intended meaning" when referring to literature. When referring to non-literary elements, such as a graphic on a web page, it refers to expectation in general: Incongruity between what might be expected and what actually occurs. In other words, not necessarily a matter of intention, but necessarily a matter of violating expectation.

It doesn't matter anyway, the entire FX series...

[Anonymous Coward]

... of video cards suck in UT2003. Slowest freakin framerates ever and the picture quality in both Windws 2D apps and most 3D games just looks 'dirty'. Even an old GF2 MX200 32MB card blows away the shiny new 128MB FX 5200 card I just wasted my money on.

Here's the solution to the problem...

[BeNude]

/* 3D benchmark utility */ /* pseudocode */

void main (argv,argc)
{
char argv[];
int argc;

gen_random_filename(argv[0]));

benchmark_suite();
}

Re:Here's the solution to the problem...

[Anonymous Coward]

The strings pointed to by argv[n] can't be counted on to be writable. Might work on some platforms, but you can't count on it. Of course, what you really do is have a little utility that renames the executable and runs that.

Re:Here's the solution to the problem...

[Anonymous Coward]

errr, that's:

char *argv[]

ATI/Nvidia cheat, FutureMark is spineless

[T5]

Forget about Intel vs. AMD, RDRAM vs. DDR. This is the real political intrigue now. Two cheating hardware companies and the benchmark tool company who hasn't got the guts to stand behind the truth of the matter.

Once again, and this can't be stated strongly enough - synthetic benchmarks really don't tell you what you think you're hearing. Indicative? Yes. Conclusive? Absolutely not. Don't listen too deeply to them.

When this much money is at stake, don't expect to hear the truth from any angle associated with these companies. Remember, we're dealing with marketers and lawyers here...

Doesn't really matter...

[mraymer]

As some people on this thread have pointed out, if you're the type of person that'll plunk a few hundred on a video card based on its 3dmark scores, you have more to worry about that being out a few hundred dollars. Heheh...

Most major vendors have been "optimizing" for Quake III ever since it became the informal benchmarking standard... I think Futuremark has blown the issue up a little since they weren't on really good terms with nVidia before this started.

But really, 3dmark has always been a "gee-whiz" pretty demo of current graphic card abilities, but never a reliable benchmark. In fact, no one program/game can be a reliable benchmark, since performance must be judged on a variety of applications. Only then do you get some kind of idea of where the "real world" performance lays.

The competition between ATI and nVidia is good for us customers; they both have excellent cards now. ATI has the fastest, while nVidia's drivers (yeah linux support is flakey I know) seem a bit more stable than ATIâ(TM)s.

Really, between the two companies, it is hard to make a really "wrong" choice.

So yeah, everyone, these aren't the droids you are looking for, you can go about your business...

Re:Doesn't really matter...

[evilviper]

(yeah linux support is flakey I know)

Be thankful that there is Linux support at all. With any other company, you have vendor support, drivers written by a volunteer third party who most likely had to reverse-engineer everything, etc. You will NEVER get TV-out working on an ATI card under Linux for instance, and TV-in isn't terribly functional either.

Really, between the two companies, it is hard to make a really "wrong" choice.

That would be true if the hardware was the only issue. Since the software is important to make the hardware work, stick with Nvidia if you need to use the card on Linux/FreeBSD.

Or, better yet, piss them both off (as well as whatever store you are shopping in) and, instead of spending $200 on the latest and greatest, spend $20 on a SIS card that's nearly as good. :-) SIS drivers were reverse engineered, but they work quite well anyhow. Send off an angry letter to SIS while you are at it, and tell them to open their docs.

Re:Doesn't really matter...

[mraymer]

I would just like to say I totally agree with you, but I had to add in the flakey linux support thing, because the last time I didn't I was flamed to hell. ;)

Last time all I heard was "blah blah the driver is closed blah blah it crashed my linux box" so I am careful now, hehe.

Re:Doesn't really matter...

[evilviper]

Well, some of it is provided binary-only (LibGL, libXv), but most is provided in source-code form. I believe it would be legal to publically release patches to the source (non-unified diff, so it doesn't contain any of the original code)

In that way, the Nvidia driver is no more restrictive than all of DJB's utilities.

Hypocrisy

[Hamfist]

People don't complain when the latest driver rev improves their Quake III performance.

Quake III is a much more respected benchmark than 3DMark, as it's actually a game.

Both programs are used as benchmarks.

If anything it tells me that NVidia engineers can optimize well for specific programs, which is part of their job. This is a good thing if I am comparing one driver rev to the next and deciding whether or not to use the new driver.

I don't see what the fuss is all about, unless one enjoys sitting watching becnhmarks run.

Re:Hypocrisy

[EmagGeek]

Well I was going to reply to this, and spent 10 minutes typing examples, but their fucking "lameness filter" wouldn't let me write " open-parenthesis-capital-x-closed-parenthesis" and I got pissed off..

Re:Hypocrisy

[Anonymous Coward]

Frankly, for what it's worth, I don't normally view "application specific optimizations" as being cheats. All the better if a company is willing to devote resources to improve the performance on specific titles. However, there is real value in having an unbiased, unoptimized benchmark so that people who develop applications that aren't privvy to these optimizations have a metric for knowing which card will likely run their applications better. Not everyone writes a 3D application with a commitment to use vendor-specific extensions/paths or with an expectation that their application is going to be "optimized" at the driver level.

ATI got burned years ago for "cheating" and they still have to live with the consequences today. Now nVidia is caught "cheating" and it seems an entire community is willing to brush it off. Fact is, both companies cheated with FutureMark's 3DMark03---though one could argue ATI less so than nVidia---but ATI is at least willingly backing out their "optimizations" while nVidia refuses and is, obviously, using its legal weight to bully FutureMark.

Re:Hypocrisy

[Anonymous Coward]

Quake 3 is a benchmark of what? How video cards perform on outdated game technology?

What good is a benchmark that doesn't test anything new? Unless you plan to spend most of your time playing it, using Quake 3 as a benchmark is completely worthless.

Re:Hypocrisy

[minus9]

If anything it tells me that NVidia engineers can optimize well for specific programs, which is part of their job. This is a good thing if I am comparing one driver rev to the next and deciding whether or not to use the new driver.

But shouldn't they be optimizing for OpenGL and DirectX rather than specific programs?

Re:Hypocrisy

[Hamfist]

They should be doing both.

Optimizing a specific program that has a million or more users is very worthwhile, as application-specific optimizations can often have greater impact on framerate or visual quality than API optimizations.

id software generally convinces Nvidia and ATI to optimize for their program before it's even released.

With graphics programming moving in shader languages, the picture changs again.

If a popular game is using a standard DirectX 9 Shader, optimizing that shader for the way that application uses the shader in turn optimizes that program.

Many 'cheats' from 3DMark are of that type. Optimizing a particular shader for the way a specific application uses it. If the base spec of a shader requires a clear of a buffer at a certain point, but the specific application does not require it because of how it manages the scene, you can improve performance for that program by optimizing the shader for the program.

The Doom 3 speedups with NVidia are becuase they optimized some specific ARB extensions. ARB extensions proposed by NVidia.

The only thing disappointing about the 3DMark optimizations is the swapped speed for quality. That seems like a bit lame, as many users would prefer quality over 2fps more.

The best way to do all this trasparently would be the ability to enable or disable specific optimizacions in the driver. Explicitly stating what your optimizations are is much safer than letting perople discover them, as Nvidia is finding out.

Re:Hypocrisy

[Performer Guy]

Anyone can optimize if it means degrading quality, the real hypocricy is degrading quality and not doing what a benchmark is trying to mention, while boasting about superior quality rendering. It was NVIDIA saying "hey we do high quality 8X aniso" when they didn't. Any fool can get that performance benefit if they want, without some NVIDIA hack lying to them. Just go to display->properties->advanced->D3D and crank down your iniso filtering to override the application's settings, voilla, so don't pretend this was an "optimization". It was a cheat. The fundamental point here is they were telling reviewers they were doing 8X aniso when some driver engineer had hotwired it under the covers do NOT do 8X anisotropic filtering. That's fraudulent where I come from.

Re:Hypocrisy

[legion88]

Yes, they should be optimizing for the API (OGL or DX) instead of specific programs. But they won't do it when there's an opportunity to make their hardware look faster in a specific application. ATi did it before with 3DMark2001 (http://www.legion88.org). And many of us still remember ATi's Quake 3 benchmark scandal (i.e. "Quack"). ATi didn't actually suffer anything from that even though people found out about it. I guess NVIDIA thought they could have gotten away with it, too.

Re:Hypocrisy

[Hamfist]

If that's the case, then perhaps I should've read a bit more before opening my big mouth....

I was under the impression that they had only optimized a specific 'standard' DirectX 9 Shader for the way that 3DMark used it. Do you have a hyperlink or anything with your more detailed information?

Please!

[Komarosu]

i say please include application based optmizations at will, just don't do it for benchmarking tools...i would like my games to run just that little bit faster, imho i take no notice to benchmarks just real word tests.

My Gforce4 is enough to run any games i want, and any future games i cant run then i upgrade to something the developer recommends...

Re:Please!

[MidoriKid]

Instead of begging for hardware manufacturers to force you to turn off better looking texture filtering or antialiasing, why don't you just go into the game options and lower the graphic detail yourself?

Or did you think these "application specific optimizations" did something else? I didn't actually think someone would have fallen for the marketing spin on this trend. *sigh*

Re:Please!

[Komarosu]

nah im saying if the application uses a certain method of rendering i see it better for em to improve the code a slight bit to make that method run a little faster?

Also read some other comments, it says they improve quality in the benchmark? I'm not listening to the market spin, i just think Nvida could help tweak applications at the driver level if they just put some effort in, and give some positive gains that people will see.

Card change

[S.I.O.]

Dang! I just replaced my GeForge MX with an ATI 9700 - now I have to rename all of my EXEs from 3dmark03.exe to quake.exe

This is about trust and users

[the-banker]

Users want to know how a piece of hardware performs. When a hardware vendor takes a shortcut it improve results against a specific benchmark, it is subverting the purpose of the benchmark and is unethical. By 'optimizing' or 'cheating', NVIDIA simply has created a situation where the benchmark is not indicative of real-world performance, and consumers lose a source of factual data.

It would not surprise me to see that much of this is an attempt by NVIDIA to marginalize the value of FutureMark 3d 2003. If a benchmark isn't favorable to a piece of hardware, then make the benchmark a 3-ring circus with these antics - then nobody trusts the benchmark at all.

A sad way to do business and I can't say when my GF3 Ti 200 will be replaced, but it when it is I will not be using NVIDIA. Apparently they don't trust users to make a decision based on an honest assessment of facts.

Wrong

[ionpro]

Doom 3 uses extensive DX9 features, such as stencil shadowing and shader programs. NO cards prior to NV3x and R3xx had those features. Sure, you may be able to PLAY Doom 3 on a lesser card, but you certainly won't be getting the full experience (or more then a few frames per second, most likely)

Re:Wrong

[Anonymous Coward]

Not true, Doom3 is actually based on the Geforce1 feature set. Hardware T+L, per pixel bumpmapping and stencil shadows are the core of the engine. Also it uses OpenGl not DX9. There are shader programs there, but they are all possible using the fixed functions present in DX7 hardware, and register combiners.

Now there could be a problem with fillrate on older cards, as Doom3 uses many passes to get the lighting and shadows calculated. But remember 640*480 only needs 1/4 the fillrate of 1280*960, so just keep the resolution low.

Re:Wrong

[10Ghz]

Doom3 uses OpenGL, not Direct3D, so it does not have "DX9 features". And fact is that features-wise, it's targeted at GeForce, not GeForce FX. Of course FX will run it faster, but original GeForce has the needed features to run D3.

Image Quality

[Jade E. 2]

OK, so changing the filtering settings based on executable name is bad, we all agree on that, but has anyone really looked at the screenshots they posted?

I opened the optimized screen and the renamed .exe screen in two tabs and have been scrolling to each corner flipping back and forth between them, and I've got to say I actually think the image quality is higher on the optimized one. If you look at the book in the background, you get a hint of text on it in the optimized version, where it's blank in the renamed one. And the bevels on the edges of the desks are a lot clearer in the optimized one. And there's a really jagged edge on the carpet under the left desk in the renamed one that gets fixed in the optimized one. It's not all good, of course. There are some textures on the left wall that are brighter in the renamed one, but it's hard to tell which one would be better without seeing it in action. (My system gets about 10fps on that test, not really enough.) And finally the optimized one has one thing that looks obviously worse, and that's the cross pieces on the rear window, they're a little strange.

So, anyways, even though it's bad that they change the settings to get higher scores on the benchmark, I'd like to know how to change those settings myself, if it improves performance that much and looks (arguably) no worse, or even better.

Re:Image Quality

[ionpro]

Right. So we should congratuate nVidia for producing a card that has worse image quality when not specifically optimized for a application? Doesn't this show that the default optimizations are, in fact, already reducing image quality? For more comparison, I suggest you go to any one of the reviews of R9700 vs. FX5800 and check the image quality comparisons versus the DX9 reference software renderer.

Re:Image Quality

[Anonymous Coward]

That's because you've got them confused. The "optimized" one is 3dmark, the renamed one is 3dmurk . 3dmurk has the smooth carpet and the hints of text on the book. 3dmark has the jaggies.

Re:Image Quality

[jdh-22]

There should be no visiable differance. Take a look at their next screen shot where they compare the two images, and you can see that there is texture filtering being used. It helps to read the article.

Re:Image Quality

[evilviper]

I actually think the image quality is higher on the optimized one.

So, to make you happy, I will build a videocard that, when it detects a certain filename, instead of rendering the picture, it will play a video of a high quality rendering of the picture... That way, it looks great, seems fast, and doesn't need to bother with doing the actual WORK.

More seriously though, the unoptomized one is much sharper. You might think it looks better because it blurs some of what would be jagged edges, but for that, you loose lots of details. Take a good look at the top of the desk. In the optomized one, the desk looks like it is flat, when it the unoptomized one you can see that it has another level on top of it (there is another rectangle drawn that doesn't show up when optomized). I guess the picture just isn't the best example. I'm sure someone could make a much better example picture, where fine details from a face or something similar will look bad when optimized.

It's really besides the point though. The issue is NVidia intentionally decieving the public.

Re:Image Quality

[Lightman_73]

So, to make you happy, I will build a videocard that, when it detects a certain filename, instead of rendering the picture, it will play a video of a high quality rendering of the picture... That way, it looks great, seems fast, and doesn't need to bother with doing the actual WORK.

No, it ain't the same thing. If a find a way to render the same scene at a faster rate, with minimal quality difference, it's not the same thing as if I would be to embed a mpg of the scene in the drivers.

I guess the picture just isn't the best example.

Here, I have to agree ;).

I'm sure someone could make a much better example picture, where fine details from a face or something similar will look bad when optimized.

Here, I can't agree ;).

How can you be so sure that there indeed is some part of the benchmark were the quality is indeed so much worse ? Should there be, I think Tech-Report would have used it, instead of the dark and not-so-good-example image they used.

Re:Image Quality

[evilviper]

No, it ain't the same thing.

No it isn't. I was just pointing out that a videocard company acting sleezy isn't a good thing. Just because they got something to go faster, doesn't mean something good is happing there.

How can you be so sure that there indeed is some part of the benchmark were the quality is indeed so much worse

Not necessarily in the benchmark. You were speaking of using this for other applications, so I was doing the same. I'm sure there would be plenty of instances where their bluring of details would be quite a bad thing, producing a lowsy picture.

Re:Image Quality

[evilviper]

You were speaking of ...

That was supposed to mean "the parent was speaking of..." Lack of sleep does fun things like that...

How about this

[confusion]

How about optimizing the driver for all applications???

Or is that just being silly?

Then can somebody explain...

[mwood]

...what is the difference between "optimizing for a benchmark" and "cheating"? I don't see it.

Re:Then can somebody explain...

[Anonymous Coward]

Optimizing doesn't degrade quality, it's a way to see look closely at what "this" application does and how "this" application can be made to run better with some changes from the standard code.

Cheating is fooling people to belive they are getting one quality setting while the driver uses another worse one.

Re:Then can somebody explain...

[jjhlk]

Cheating produces a different image, or renders a different amount of a level [that a game wouldn't be able to figure out dynamically].

Optimizing for a benchmark could still be considered cheating, since the company replaces a shader with a custom one. Even if all it does only take advantage of video card architecture, I don't know if it warrants being allowed to do somethiong like that.

Press release from NVidia

[DaveOke]

I now inform you that you are too far from reality. There are no cheats in our drivers. Never! I blame Tech-net - they are marketing for ATI! ATI, they always depend on a method what I call ... stupid, silly. We are not afraid of ATI. 3DMurk has condemned them. They are stupid. They are stupid.......and they are condemned. All I ask is check yourself. Do not in fact repeat their lies. I have detailed information about the situation...which completely proves that what they allege are illusions... They lie every day. I triple guarantee you, there are no cheats in our drivers.

Re:Press release from NVidia

[furmac]

Your just upset cause your ATI stock keeps going down :)

Re:Press release from NVidia

[damiam]

Actually, ATI stock is doing pretty well.

Re:Press release from NVidia

[Anonymous Coward]

This is a lie. There are no ATI soldiers in our compound. Allah be praised! We will set the ATI infidels on fire and butcher their women and children in the street like the pigs they are!

NVidia loves the Iraqi Information Minister too. :)

Wheeeee!!!

[Anonymous Coward]

nVidia rocks, ATI socks. That's the way I see it. Who would have thought that so much drama whould be created over a company's driver optimizations??! It's kind of funny if you step back and think about it.

Benchmarks are useless

[Dr. Bent]

When I want to test-drive a new car, do I build a driving simulator based on the car and load in my route to work so I can see how it would handle during a typical commute? Of course not, I just take the car for a test drive, because sometimes the best way to test something is just to use it.

Benchmarks are inheriently flawed because a benchmark is not what your buy the card for. Card manufacturers are always going to "teach to the test" by optimizing their cards and drivers for whatever the reviewers are using to review the cards. So why not take advantage of this and use popular applications to benchmark? The venerable Quake3 framerate test is one example of this, but I would arge that all benchmarks should be numerical data taken from the performance of real world applications. That way, at least those apps will perform as advertised, and any other apps that take advantage of the same features on the card will benefit.

Re:Benchmarks are useless

[Anonymous Coward]

If you buy a terrain vehicle for a desert expedition, and you happen to live far away from the desert. Then you might try driving it on a test track to get a feel of how it handles in a simulated desert environment.

If you then go to the desert and discover that the car has been optimised for the test track and not the real desert you would get pretty mad (Tires that can only handle cold sand, shocks that only last the first hundred miles, cooling system that overheats after one hour use and so on) That is what NVidia has done here.

Re:Benchmarks are useless

[davFr]

And what do you do when you buy a new VCR? You ask to have it "just for the evening"? And so on with every product one can buy! Benchmarks and reviews are important and cheating IS BAD. Anyway, doping in sport must be ok then? Well we all know sportmen are cheating, thus what's the worth with those damn rules!;o) Cheating rulez! Truth suxx!

Re:Benchmarks are useless

[El_Ge_Ex]

Do you seriously open your mouth before hearing what everyone has to say!?! If I want a graphics card that performs well on technology that was state-of-the-art two years ago, I'll use current games. Then in two months it will become obsolite and I'll go buy a new one. Does that sound like a smart decision? It is what after all you are recommending!

You have to realize those games started development up to two years ago and use graphics card technology from that time (Direct X 7 anyone?) Now we are at Direct X 9, with vertex and pixel shaders!!!

You may not realize the importance of this, the the graphics performance of these cards now matter more than ever thanx to these new programmable pipelines. With these cards developers now have freedom in real-time that was only possible with SGI's renderman in the past (and NOT in real-time).

Personally, I realized when the 8500 was caught on the Quark3 issue that ATI was still a better deal and bought the card anyways. ATI has been better on the "bang-for-the-buck" benchmark since the Rage 128 and I don't expect that to change, even when they're #1. :)

Re:Benchmarks are useless

[mr3038]

When I want to test-drive a new car, do I build a driving simulator based on the car and load in my route to work so I can see how it would handle during a typical commute? Of course not, I just take the car for a test drive, because sometimes the best way to test something is just to use it.

Sure, and that's how hard drive benchmarks are done. With newest graphics cards it's the other way around: you have the car but you have no road to run on with it.

How do you test a train that requires new kind of track? You build a test track and try to guess how the train performs in real use after testing it on that test track. It's the same thing here.

Re:Benchmarks are useless

[Anonymous Coward]

You do realize that a test drive is a benchmark? Test drives usually are just within a few miles of the dealership and not for very long. Sure you can take it one the highway for a few miles and on surface streets for a little while, but is that the exact places and times you're going to use the car for?

No, these aren't the places and times you're going to use that car. In other words, you aren't testing it against the game you'll be playing it in. But I also don't know anyone that wouldn't take a car for a test drive before they bought it, so there must be some value there in making a short drive over a roads you probably never drive regularly.

And how do you narrow down your car choices to begin with? Do you drive every model of every car available for purchase? No, you narrow down your parameters using, get this, outside sources. Maybe you buy a Car and Driver or a Consumer Reports to see how the cars perform, or maybe you just pay attention to the manufacturers specs.

NVIDIA is basically saying that their car performs amazingly well, as long as the course only contains left turns. Or as long as the course is perfectly smooth. Or more realistically, as long as they can examine the course closely before hand and program the car's active suspension to the course's exact layout.

Benchmarks are everywhere. They are useful, especially for narrowing down your choices.

why 3dmark is irrelevant for many gamers

[Scudsucker]

Most of the pro-3dmark comments are pointing out that it tests future technology that today's games don't use. That's irrelevant because for most gamers because by the time you actually need those features you need to upgrade your card anyway! Look at Doom 3: my TNT2 card can't play it because the TNT's don't have hardware T&L. But even if the card did support it, I'd need a new card to play the game anyway.

I could see people making arguments that DX9 tests are good because DX9 games will prolly start showing up in the next year or so. But I'm not too worried about Nvidia's DX9's performance as they've been partnered with Microsoft to manufacture the XBox.

Re:why 3dmark is irrelevant for many gamers

[Anonymous Coward]

and XBox uses DX9? oh wait, you wanted to imply that the Nvidia-MS cooperation on Xbox spilled over to excellent DX9 performance on NV's newest cards. guess what, it hasn't materialized. if you'd have paid attention to individual testscores in 3DMark03, you'd have known that. your loss.
not only that, DX9 was developed and tested on ATI's prototype R300 cards. I believe this would be a much better indicator of expected DX9 performance.

Re:why 3dmark is irrelevant for many gamers

[Anonymous Coward]

if you'd have paid attention to individual testscores in 3DMark03, you'd have known that. your loss.

If you had been paying attention for the last month, you'd know that any 3DMark scores are suspect one way or the other. But you probably missed that obvious point in your rush to be a condesending Anonymous Cockgobbler.

Re:So What? Who Cares?

[TheDredd]

But if you can't notice the difference by eye, who cares They should make this a feature in the driver settings for the end user, so you can increase your framerate with your favourite game

pfth give me a tseng et2000 card

[Openadvocate]

You can talk all you want about your fancy graphic card, but I'll stick to my tseng et-2000.

Re:So What? Who Cares?

[Coyote67]

To be fair, it is the /. crowd that 3dmark appeals to. 3dmark has never been more than a simulation of a game and as such should be considered useless. Games that people actually play are the only real benchmarks as they represent real-world results, the only results that matter. Nvidia optimizes it's drivers for games(benchmarks) and so does every other 3D video card manufacturer, is that cheating? No thats improving performance by tailoring the drivers for a specific enviroment. Thats optimization. Is compiling for athlons cheating? Course its not.
The only cheating is on futuremark's part for selling a product that they claim is a valid benchmark for 3d gaming when it is nothing of the sort.

this is all cockwaving

[Anonymous Coward]

The really great part is that the two cards taht they're bickering over have such ungodly good performance that:

1. The human eye will not consciously notice the difference in quality or performance, on average.
2. The FPS on the cards is so much further above typical monitor refresh rates that your resource limit is going to be on the display side.
3. This entire thing boils down to ammunition for fat ass introverted slashdot idiots to debate back and forth over whose peni^H^H^H video card is bigger.

And yes, I also think people who do the ^H^H^H are not clever and betraying their hopeless nerdosity. I am mocking you.

Re:this is all cockwaving

[eatdave13]

OH MY GOD PLEASE STOP MOCKING ME!

x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x.x

Re:So What? Who Cares?

[minus9]

My Radeon 9700 pro works fine for me.

Pretty much any graphics card has drivers for XFree86, the rest should work with the vesa driver.

125 fps?

[boomka]

that sounds silly, you can't possibly need 125 fps, most
monitors only update the picture with a frequency
of 80Hz or so, so 1/3 of your frames will just never be drawn.
Anything above 80fps is just useless.

Re:125 fps?

[Kynde]

that sounds silly, you can't possibly need 125 fps, most monitors only update the picture with a frequency of 80Hz or so, so 1/3 of your frames will just never be drawn.
Anything above 80fps is just useless.

If this is about Quake 3 again, the 125 fps is not about graphics, but physics. Exactly 125 (and 43 and 76 IIRC) maximises the rounding errors that occur in player physics simulation, this allow people to pull off a tad higher rocket jumps and other such trickery. This was an accident, but it's being widely used by using com_maxfps to cap the fps to either 76 or 125 and making sure your 3D card can sustain that in all situations.

Whats up?

[Anonymous Coward]

(X)

Why not allow the user to choose?

[MrWorf]

I mean, in this age were we are supposed to tweak the bejesus out of everything (just look at all the people with computers in refridgerators :) ), why not simply let the users choose themselfs?

Just add a new page in the driver settings where you can add an exe-file and then allow the user to activate the different "tweaks/optimizations". It would be more honest and people who want more speed than looks would be happy as a clam.

"But won't the user be confused by all the options?" I hear you ask. Maybe, but have you looked at the BIOS setup lately? Now, the wrong settings here can potentially blow the socks of your precious processor, but the wrong tweak setting will only make DOOM XII look ugly.

Anyway, thats my five cents on this issue. Enough with this and onwards to new drivers and more piethrowing ;)

cheats vs. optimizations

[slyguy420]

I read the original Extremetech article which details them using the beta version of 3dmark2k3 to "stop" the demo and move the camera outside of the normal rendered path which revealed that nvidia played with their drivers to take a load off the card by not fully renedering everything seen "outside" of the normal view. I would no consider this a cheat, but an optimization. who cares what is not seen by the camera? I give nvidia a pat on the back for this.

On the other hand when you enable the 8x aniso filtering and the driver deliberately reduces image quality to gain a few more marks, I would consider this unfair cheating. If I ask for 8x aniso you better damn well give it to me.

Lastly I think it's stupid for anyone to rely on any single benchmarking program to gauge the performance of a card, that's why I always read THG articles when I want to know how hardware performs, Tom uses 3dmark as one of many applications that he benchmarks new hardware with. and frankly I don't really care about 3dmark scores, I always look at the scores for real games.. I wanna see how my gaming is going to be affected, not how sum st00pid benchmark runs.

Re:cheats vs. optimizations

[untaken_name]

I read the original Extremetech article which details them using the beta version of 3dmark2k3 to "stop" the demo and move the camera outside of the normal rendered path which revealed that nvidia played with their drivers to take a load off the card by not fully renedering everything seen "outside" of the normal view. I would no consider this a cheat, but an optimization. who cares what is not seen by the camera? I give nvidia a pat on the back for this.

You're quite missing the point. nvidia decided to optimize the static path the camera takes through that particular demo. Now, if all games were just static movies, this would be optimization. However, the demo was supposed to simulate an *engine*, not just a pre-determined camera path. Nvidia *only* optimized the textures *on that path*, and in fact made other camera paths look distorted. ATIs cards, on the other hand, had the same quality no matter *which* camera path was taken. This is why it was cheating, not optimization. Optimization would have made *any* camera path run better, not just the default one. What good does it do to 'optimize' only one narrow camera path? I know I'd prefer to be able to look around, maybe move in a direction *I* chose, within a game. This is the equivalent of a pre-rendered cutscene using the 'in game engine' rather than a real-time rendering of the same engine. I don't give anyone a pat on the back for openly cheating.

Renaming my q3.exe

[Japie]

If I rename my quake3's q3.exe to 3DMark03.exe will I get some extra fps? :-)

Ok ok, I shut up....

Thing that always makes me laugh

[Anonymous Coward]

Is when people by high end video cards (read: the latest offerings from nVidia and ATI) and use a fricking 15" display even 17" sometimes.

These cards are meant for games to run 1600x1200 @ 85 hz or higher with AA and AF on.

I know people who buy 3 GHz chips, GB of RAM, Radeon 9800 Pro and re use a blurry 15" 1024x768 wonder.

Wrong.

[Anonymous Coward]

The reference designs do NOT specify the type and size of fan (or other implement) to be used on the chipset.

They only specify the cooling requirements for the chip - leaving the implementation up to third party companies. They could ship the cards with peltiers if they wanted. Or, they could severely underclock the chips to minimize heat production. Of course, that would be ridiculous, but those choises are -entirely- up to the third party vendors.

Re:So What? Who Cares?

[Slack-Jawed Local]

Absolutely! I can't believe that some people are saying this isn't cheating!! Here's the deal, prior to any optimisation/cheating... App programmer tells card to render something -> Card renders it the way app programmer tells it to -> User sees what app programmer intended. After optimisation... App programmer tells card to render something -> Card renders it the way app programmer tells it to, but faster -> User sees what app programmer intended. After cheating... Programmer tells card to render something -> Driver programmer decides that, actually app programmer doesn't know what he/she is talking about and shouldn't have told it to render the thing that way and that they know a much better way to render it -> Card renders it the way the driver programmer tells it to, which (surprise!) is faster -> User sees what driver programmer intended. The point is that what the app programmer and driver programmer intended are different things. This, in itself is not a cheat. The cheat comes in when the driver programmer doesn't tell people about the change and instead let's people think that a difference in FPS between competitor cards is because of differences in power, rather than differences in what they are trying to render. It's a matter of trust that graphics cards render things the way the app tells them to. To do otherwise is cheating. Plain and simple.

Insightful?

[cgenman]

This is not an nVidia problem? This is a benchmark problem, because they weren't wily enough to prevent nVidia from cheating tremendously and repeatedly?

People are crying about this because they rely upon benchmarks as a gague of how powerful a card is, and make purchasing decisions around such knowledge. Sure, some of them forget that a %5 difference is meaningless in real-world performance, but that doesn't mean that the overall scores are meaningless. nVidia's last round of cheats pretended that the card was 25% faster overall than it actually was. This particular cheat ads between 8 and 18 percent to the total, with the largest false total going to the most expensive card.

In other words, if you bought a $300 nVidia card on the strength of this benchmark, you bought a card that is %40 slower than it should be because nVidia went out of its way to lie about the speed.

ATI optimized for the test, they re-ordered the way in which the card handled executions similar to the way someone might re-order their day for maximum efficiency. It was a cheat, but a minor one that only added 2% to the score. nVidia's cheat involved dropping instructions entirely, equivalent to doing more in a day by checking things off your list without actually doing them, letting the food rot in the kitchen and the dirty laundry pile up.

The sad fact of the matter is that nVidia now has a big problem, in that their fastest card which managed to eeek out ATI's fastest cards can no longer claim that crown, and yet it is ATI's turn next to introduce faster cards. Their technology can compete, but can't demand the premium that graphics card developers rely upon to survive. Furthermore, this cheat comes after nVidia promised to clean up their act and remove all cheats from their driver. Not only did they cheat, they promised to clean up their act and yet cheated again in the very driver that is supposed to be clean. Their public image is bloody shot, significantly worse than ATI's was over their Quake 3 debackle. ATI fell back on their technology and released superior cards, but nVidia doesn't seem to be able to head down that road.

To get back to the poster's original position, the benchmark should try to outsmart the developers, though being a test for future games they have to stay abreast of display technology more than cheating techniques. But blaming the benchmark not the nVidia for cheating is like discovering that during a Car-and-Driver top-test the Ford team took a shortcut shaving %40 off of the race course and congratulating them on their ingenuity.

What they did was indefensible. They knew it, they appologized, and they did it again.

Re:Insightful?

[LilMikey]

"nVidia's cheat involved dropping instructions entirely, equivalent to doing more in a day by checking things off your list without actually doing them, letting the food rot in the kitchen and the dirty laundry pile up."

I never do laundry and 75% of my fridge is rotten. I guess I'm cheating life! Actually, those things were never on my list so I guess it doesn't count.

Re:Insightful?

[EriDay]

This is a benchmark problem, because they weren't wily enough to prevent nVidia from cheating tremendously and repeatedly?

Repeatedly.

Fool me once, shame on you, fool me twice, shame on me.

I'm not saying what nVidia did was right. My point is that with the millions of $$$ are at stake, why isn't a benchmark smart enough to detect this used. If it were my benchmark, I'd give them a 0 with an asterisk stating that they had cheated. A few 0's and they'd either mend their ways or go out of business.

Re:Insightful?

[trashme]

I'm not saying what nVidia did was right. My point is that with the millions of $$$ are at stake, why isn't a benchmark smart enough to detect this used.

Now you're talking about an arms race between the benchmark and the driver/hardware. Should Futuremark be putting their resources into trying to create a fair benchmark that is representative of the real world use of a video card? Or should they be tossing extra code into their benchmark that tries to detect cheats? Even if they are able to detect this cheat, what about the next one somebody comes up with?

The point is what nVidia did is dishonest. I don't blame the benchmark. I blame the company that intentionally inflates their benchmark scores and lies to the consumer.

Re:Insightful?

[EriDay]

Their role is to tell us which is the good stuff. Anderson's auditing firm couldn't tell us which was the good stuff and they went out of business.

The 1st time it happened, they should have been shocked. The 2nd time it happened, they should have been prepared.

Re:Insightful?

[jjhlk]

Why bother to try and find cheats when this is merely a benchmark of DX9 performance? Everyone knows, you matter how hard you try, it's impossible to secure something. (Half-Life is an old example)

Re:Insightful?

[menasius]

I think you are too quick to say that ATI should have a better Reputation than nVidia. Both companies have rampant use of these "optimizations".

nVidia likes the name 3Dmark2003
ATI likes the name Quake3

nVidia is dropping instructions
ATI is dropping bits (whats that... no 32 bit precision, only 24? not quite what ARB_pixel_shader meant).

The fact of the matter is heads up comparison of these cards with future technologies can NOT be done anyhow. There is no equivalent render path between the two of them, especially when you are talking shaders. Difference in precision, whether explicit or implicit makes this impossible. Shockingly, no matter the "cheat" the one running at the lower precision usually wins. The 3Dmark2003 cheat actually allows a 32bit precision nVidia card to eek out the 24 bit ATI card.

-bort

Re:Insightful?

[Anonymous Coward]

the 3DMark03 cheat drops precision to FP16/INT12 on the FX cards. your claim that 32bits eek out ATI's 24 is a lie.

ARB_pixel_shader does not demand 32bits. your claim that it does is another lie.

ATI is not dropping bits. DX9 requires FP24 minimum, which ATI runs at ALL THE TIME. if you or Nvidia or ACME_3D want to go overboard with 512x10^1024 bit-precision, feel free. Just don't complain and cheat your ass off later on when you realize your uber-precision masterpiece runs like ass.

funny that you mention the ARB "standard". if nvidia were so great at excelling in ARB rendering paths, how come DoomIII has to use a custom NV30 path with reduced FP16 precision to run at acceptable performance on NV30/35 cards?

Re:Insightful?

[Nurf]

In other words, if you bought a $300 nVidia card on the strength of this benchmark, you bought a card that is %40 slower than it should be because nVidia went out of its way to lie about the speed.

Hmm. You know, this sounds like a great class action lawsuit. Even if the award wasn't too great, it would drive home the point nicely that cheating was stupid and self-destructive.

Of course, so few FX series cards have been sold that I think nVidia wouldnt have to pay out much. :-P

(Oh, and it's spelt "eke", btw. :-) I've seen a too few "eek"s lately to ignore that one. Please forgive me being anal.)

Ciao!

Re:So What? Who Cares?

[TwistedGreen]

Why's that sad? Who really cares? Most people just want a decent card, and either don't care enough or don't have time to devote themselves to some absurd quest to find the perfect graphics card. You'll never get the latest technologies in a consumer-grade product, so as long as you're getting a good price and it works, why should they care? Stop making this out to be some monstrous injustice, because it's not.

Nice Demonstration

[Anonymous Coward]

Nvidia spending extra development money to "cheat" on specific products is only a nice proof of how stupid we are: we just buy whichever cards give the best FPS, even though they're already 3 to 10 times faster than what we can see.

Mod Parent ???

[JohnnySkidmarks]

It say's nothing in post. Why is this mod'd to 5(?!) blazing points??? If nothing else, this post is off-topic. How does M$ cheat? They are a vicious Litigation-adled pool of scum bags for sure, But certainly no more so than Martha Stewart, Dick Cheney, or oh nevermind, Now I'm the Troll Right?

Checks and balances

[cgenman]

I find it interesting that it was ATI that blew the whistle on NVIDIA, and NVIDIA that blew the whistle on ATI. Personally I hope they both pour more resources into auditing the behavior of eachother's drivers, so that cheats like this can be revealed.

At least someone is checking.

Re:So What? Who Cares?

[s20451]

Maybe my main problem is that I prefer to use laptops than desktops. Yes, the desktop hardware situation has improved markedly, but my Toshiba laptop required some severe tweaking and patching when it was new, just under a year ago. Recently I installed RH9, and it was such a disaster that I had to revert to my original RH7.3.

Re:So What? Who Cares?

[Jason_says]

insert funny penny-arcade comic strip here

Screen shot

[Anonymous Coward]

This page has the damning evidence:


http://www.tech-report.com/etc/2003q2/3dmurk03/i nd ex.x?pg=6

-AC

Re:Image Quality (MOD PARENT UP)

[Lightman_73]

First of all, the differences in image quality are, as the parent already pointed out, negligible.

Thus, there are two questions that I want to ask:

First, sure, NVidia included some optimizations specifically targeted for 3DMark2003. But, the real question is, if the (perceived) image quality doesn't suffer (and it doesn't, as you all can see by looking at the two screenshots tech-report posted), is it really so terrible ? All gfx cards manufacturers did and do these kind of things. I can't see what's the problem (in this specific case) is.

Then, we have the second, and more important question. Why the hell Tech-report is so fast at pointing out these optimizations in NVidia drivers, while it hasn't (as long as I can tell) investigated so thoroughly ATI's drivers ? Note, I'm saying that ATI's cheating or whatever. I'm only pointing out the fact that afaik, ATI's drivers could change behaviour if you rename the 3DMark executable, or that they could have some specific optimizations (shader code reordening apart) for 3DMark that impact the quality as much as these NV optimizations (that is, in a negligible way).

Dunno, sounds a little strange. In particular by taking into account the FX5900 reviews at Tom's and Anand's sites. These reviews point out how the FX5900 is indeed faster than anything on the block in almost any game. And image quality isn't impacted. So.... Hrm... Dunno...

Linux Optimization

[Anonymous Coward]

I guess they should optimize the 2D Performance of their Linux drivers. My desktop is running Debian (testing since yesterday) and OpenOffice is slow as turtle. Mozilla Mail / Mozilla Firebird is the same.
And my system aint't the worst (AMD 1333MHz, 512MB, GF2MX400 64MB).

If they would optimize their drivers for Linux i would not be angered with them.

Re:Linux Optimization

[pair-a-noyd]

You should go to the nvidia website and download the Linux drivers.

Mine smokes. Q3A on my P42ghz w/GF4-Ti4200 128m is insanely fast. My son says it blows away ANY console he has ever seen and he has a PS2 and plays everything that comes out.

I'm building him a new machine this weekend, a 2.2ghz Mandrake 9.1 box w/+GF4MX440 128m for his gaming..

So far I'm very highly impressed with Nvidia preformance. I was highly disappointed and distressed with driver installation previously but they've made great improvements and it's very easy to install Nvidia drivers now..

I can't speak to the competition, haven't tried thier products yet...

Retarded

[222]

This is absolutely retarded....
3D Mark 2003 is not to be trusted, not because of some pro-Nvidia stance i have, but more so because, as its been said 100 times already, 3D Marks can often have relatively small insight as to how hardware will actually work with games. Everyone posting about how they're never going to buy another Nvidia card is just being foolish, as ATI and Nvidia will most likely both provide consistant, cutting edge hardware at a reasonable price.
It should also be taken into consideration the ammount of pressure put on these two companies, since (I still have trouble believing this) major OEM's actually use 3dmark as a factor when purchasing hardware. If i've lost respect for anyone in this whole ordeal, its Dell.
I use 3d studio max, and i play warcraft 3. On occasion, ill pop in a FPS, and I look forward to Doom ]|[. Call me silly, but performance with those applications are what govern what hardware ill be buying next, not a synthetic benchmark. ESPECIALLY one that doesnt have the source readily available...
I wouldnt mind seeing an open source GL benchmark, written by individuals that have nothing to lose or gain by a cards performance. (please, drop a link if one exists)

bah!!!

[m00by]

say a car were being benchmarked. I know, it's a strange way to say it, but hey, if you were a car maker and you "optimized" the testing area for your car on the 0-60mph test, say by sloping it down 20-30 degrees, or in a 60-0 brake test, sloping up (or perhaps even adding a brick wall =D) the track, they would definetly be accused of/crucified aboout/hung out to dry over their cheating. but hey, in computer hardware, it's just a driver optimization...

nVidia too dumb to check CRC for *.exe instead?

[Assmasher]

LOL...

Fairy.exe

[sixb0nes]

I still find it odd that renaming the infamous nVidia "Dawn Demo" from Fairy.exe to 3dmark03.exe gives you some pleasing results. The best part was watching it on my 9700 with a 3rd party gl wrapper :)

To the issue at hand, anyone who says to ignore the results of these tests because synth benchmarks mean nothing is missing the point. Who's to say your favourite game doesn't have disabled features that you have *explicity set* in a similar vain?

Re:Fairy.exe

[jericho4.0]

Yes. Odd to say the least. Why would NV use '3dmark03.exe'? Why would they then use the same mechanisim in a clumsy cheating attempt? Maybe a disgruntled employee is having a joke?

Benchmark-program names

[Salamander]

Changing the executable name should be standard procedure when running a benchmark for public consumption. Ideally, benchmark developers would go even further and actively try to cloak their identity from the drivers, databases, or whatever software the benchmark tests. You might think this could lead to an "arms race" between vendors and benchmarkers, but I don't think so. To get away with cheating, the vendors' detection of the benchmark would have to be totally foolproof. It would have to guess right every single time to avoid being caught, and that's almost as much work as making the optimization properly general in the first place. Vendors wouldn't bother, if the risk/reward ration were that poor. The benchmark "cloaking" doesn't have to be perfect any more than crypto needs to be perfect. It just needs to be good enough that folks decide another road is better.

Re:Image Quality (MOD PARENT UP)

[TheMeuge]

Actually, Tech-Report did check what happens to the ATI drivers as well, just visit their website and see. And like they said in the article, the differences they found were negligible for ATI (1-2%) and I would not be surprised if image quality remained the same. After all, there ARE legitimate optimizations one can write for a graphics engine to run better on a given video card architecture, just look at Doom3. The optimisations for GFFX are endless when it comes to the new Doom engine, and I doubt that Id would want to get caught reducing game image quality to let a certain card have an fps advantage.

P.S.I've been visiting Tech-Report for a while now and I find no reason to doubt their honesty quite yet, but then again - i semi-trusted nVidia too.

3DMark Tests Architecture?

[asdkrht]

Well, I always looked at the 3DMark benchmark as more of a way to test the architecture of the graphics controller since it doesn't have any card specific optimizations like "real world" applications would tend to have. After all, if you care how fast Doom III, HalfLife 2 or whatever, will run using a card then I imagine you would test or look at reviews where they test using those specific applications you are interested in. If the architecture is bad then that means you'll only get decent performance if the developers make the effort to add optimizations or workarounds for the limitations of the card.

Re:So What? Who Cares?

[Open_The_Box]

Fair point, but it's not the customer on the street or even the person (well informed slashdot readers, etc...) who wants high performance at a cheap price that matters in this case. What they're really bothered about it the ability to say "We've got the fastest card and here are the benchmarcks to prove it! We lead the market! Nyah!"

Or something like that anyway. Right or wrong, that's the way it is.

Re:So What? Who Cares?

[Junks Jerzey]

You are... I am... And a bunch of /. readers are too. But the millions of teenagers that grew up without technical skills but love games and subscribe to gaming magazines are looking at the benchmarks to decide what to ask for X-Mas and what to beg Mom to buy at CompUSA.

I don't know, I think it's more widespread than that. It's the same kind of fervor that surrounds arguments about the # of instruction units on the Pentium 4 vs. the Opteron. They're people with some technical knowledge--maybe even a good amount--but it's knowledge that exists in a void, without any kind of real world grounding.

Perhaps the oddest thing about the 3D video market on the PC is that the high-end is very much a niche. 95% all games released do absolutely nothing that requires a GeForce 3 or better, and the simple reason is that the vast majority of PCs sold come with lesser cards. Most Dells ship with GeForce 4s, which don't have programmable shaders. The new, slimline Dell desktop (very nice in most respects) ships with Intel's motherboard 3D, which is a couple of generations back (before hardware T&L). So developers just have to ignore most the cool, cutting edge stuff, because if they target it then they're reducing their market substantially. Remember, games like Quake 3 sold less than 200,000 copies as it is. Now on the Xbox, though, shaders are king.

Re:So What? Who Cares?

[Open_The_Box]

Sorry that should be "benchmarks". Must use preview. Bad poster. No cookie.

Re:So What? Who Cares?

[be-fan]

Um, optimization doesn't necessarily mean that there is no user-perceptable difference. In graphics, one of the main optimization techniques is to degrade quality in ways that you hope the user doesn't notice. If the user does notice, than you get a bad rep for quality, but its not cheating per se.

Re:I will take a wait and see attitude. [OT]

[mr3038]

The DualShock2 is the best controller out there for just about everything else. It's comfortable, it has intelligently placed buttons, it has enough buttons, but not *too* many

It might be the best, but it's far from good. I personally prefer DualShock2 analog sticks over other products but DualShock2 has really bad buttons when compared to competitors. Have you ever tried triggers under the XBox controller? You just hate to use DualShock2 controller ever again.

One thing I especially hate with game console controllers is that those're designed to be used with thumbs only. Hello? See, I've four other fingers in both hands, why not use those? (Yep, some games support triggers but still using all 10 buttons in the DualShock controller is real pain in the ass.)

DualShock2 (and any controller that comes with any game console) is something that allows you to play all games but it's not perferct for anything. Take driving games for example: the only driving wheel that works even moderately well with PS2 is Logitech's newer model. AFAIK, there's not a single one usable force feedback wheel controller for XBox. Nevermind the fact that the MS FF Wheel for PC is about the best wheel there is. And this is the situation even though driving games is one of the biggest console gaming genres.

And all the first person shooters... anybody that claims that console controller is more suitable for that genre of games than keyboard and mouse should go online with some game that allows plaing against PC players and get their feet wet.

The problem with all the consoles is lowest common dominator: if something doesn't come with every console it's supported by practically no game. How many games support, for example, more than 2 players on PS2 even though the console can support 8?

How I make MY decision on Video Cards...

[Frobozz0]

I make my decision on what video card to purchase based on only a couple of factors. The least inportant to me, accross same-generation cards by nVidia and ATi, is frame rate. It's ben said a thousand times that you can't see the diff. past about 60 fps if your card can maintain at least 60 fps at all times. If my card can handle all the features I want in future games, like FSAA and per-pixel shading, then it meets my criteria for feature set.

In order of preference (because so many things are similar these days):

1) I favor ATi over nVidia because I believe ATi is a better company.
2) Application support, which translates to the quality of the drivers. I use the same computer to do high end 3D animation and 2D photo editing as I do to play UT2003.
3) I favor the card that can handle all the features I want for the next 1.0 to 1.5 years, at which point I'm buying a new card.
4) Price, because most are so similar in price that even $50 more is not significant to me.
4) frame rate. unless the card is 20% slower at 1280 x 1024, this is insignificant to me.
5) Benchmarks? Bah. Meaningless.

Re:How I make MY decision on Video Cards...

[Nom du Keyboard]

features I want for the next 1.0 to 1.5 years, at which point I'm buying a new card

Can I have your old cards? Pretty please!!

Re:Image Quality (MOD PARENT UP)

[Lightman_73]

Duh, oops. I'm dumb, I have to apologize to Tech-Report.

Dunno why, but I thought it was T-R that first reported about NV's cheat with 3DMark2003 (the first one).

Anyway, I still don't understand all this fuss about this latest optimization. The difference in quality between the two "versions" are negligible, I reiterate. And, btw, I have to say that I kinda prefer the "optimized" one, just look at the carpet border, it's much better than in the "non-optimized" version ;) And, overall, lemme stress it, the differences are so minimal that T-R had to diff the two images and enhance the resulting image to let everyone see that some pixels were different (btw, if someone from T-R is reading this, could you plz provide an histogram of the diff image, maybe with a number_of_changed_pixels/total_pixel_count figure ?).

This alone should speak of optimization, and not cheating. In the end, if the final result is qualitatively (and not quantitatively) equal, the way they attain it shouldn't matter (as long as it isn't a jpg or mpg embedded in the drivers, obviously ;)).

Again, lemme apologize to the Tech-Report staff :)

Re:So What? Who Cares?

[dnoyeb]

Not really. I expect NVIDIA to beef up their product claims and press releases as do all companies today, though shamefully.

I do not expect outside firms to lie to me, ala Arthur Andersen. If Futuremark can not eliminate this, they will be useless, as the plainly see.

I can't imagine the relationship between FutureMark and any company who makes video cards to be peachy after the auditing scandals of late.

I think NVIDIA and ATI expect to be allowed a little play. But thats not in FutureMark's best interest.

Re:So What? Who Cares?

[spectral]

Benchmarks are meant to test card performance. Optimizing for them is wrong, ok? Optimizing for real games is ok, because there's benefit to the end user (better game play). But for synthetic benchmarks, it's just fucking cheating and lying.

Also, compiling for athlons isn't cheating, as long as 2+2 is still equal to 4, and not 3.96. The new set of issues (hell, the original set as well) involve the program not doing what it's supposed to do: Produce the best quality images. The first set found involved it completely screwing up if you went off the path, and the second set (the current one) involve lowering visual quality. ATi got slammed when they did it with quake3, nVidia deserves to take even more shit for doing it with something where the only benefit to the user is that the number at the end is higher. Thus, they're doing it to lie. ATi just did it cuz the cards at the time were crap, but it DID make a REAL game somewhat faster (though uglier).

Benchmark reliable!

[grantsellis]

I don't see why everyone feels that benchmarking NVidia's cards is unreliable. All you have to do is follow a simple formulae:

Future card's claimed performance * 0.6 = Actual performance

Tada! Now we still don't have to look at anything other than 3DMark when we're buying cards. :)

Re:Benchmark reliable!

[kosamae]

wait, but when they say that a card is twice as fast as a previous card, does that mean that its .6(2xOldCard) or .6(2x(.6)OldCard)?

Re:So What? Who Cares?

[afidel]

actually in the context of benchmarking the only thing that can be seen as an optimization and not a cheat is a tweak that results in identical but faster output, such as the change that ATI made which reordered some of the instruction for the shader program to better fit the way their card did things. Changing the visual output from what the benchmark calls for is flat out cheating, sorry no other way around it. You can claim that things aren't fair for NVidia because their card does FP32 or FP16 and that ATI only does FP24, but the DX9 spec calls for FP24 as a minimum, the fact that NVidia only has modes that fall above or below this is their own problem, it was a design decision. Besides they didn't even use the FP16 mode, they used int12 which looks much worse then even the fast but somewhat passable FP16 (which still would have been technically cheating but would have been a little more consideration when bring up their architecture argument).

Re:So What? Who Cares?

[TwistedGreen]

The real sad thing here is that we can't trust a hardware company to allow an objective test of their product's abilities. While benchmarks can only approximate the real-world usage patterns of a potential user, a good benchmark does that as accurately and comprehensively as possible. Benchmarks make sense. But nVidia evidently will not allow this sort of objective test. Do they really need to generate what essentially amounts to misinformation in order to maintain a "market lead"? Perhaps this behaviour should bring into question their product's /real/ quality of performance.

However, it is now in 3dMark's hands to prevent nVidia (or anyone else) from doing this again in the future. The credibility of their benchmarking product is at stake.

Though no matter what it's called, this is still cheating.

Re:So What? Who Cares?

[HohlerMann]

You have a problem with reading a review in a magazine and buying the product based on the review?

Stop buying NVidia crap

[Anonymous Coward]

There is only one way to make stop NVidia from cheating, stop bying their crap. (I liked them from the time I had Riva 128 in my PC, but I guess it was different company back then). I am suprised then NVidia looks at file names, I think that they will "fix" it in the next release by looking at something else (best way is to look at the code itself).
Everytime you buy anything that has NVidia logo on it you make it easier for them to cheat...

GeForce FX and GTA Vice City

[sublimespot]

I just got my GeForce FX 5200 and it has some terrible graphics corruption in certain parts of the city in Vice City. Im wondering if this is due to the driver issues

Re:GeForce FX and GTA Vice City

[cmason32]

I had the same problem with my FX 5200; the fix is real easy. Go into your display properties for the card and set the "Intellisample" settings to "application." Once I did that, all my problems disappeared.

Re:GeForce FX and GTA Vice City

[sixb0nes]

I bet they clipped the planes to gain some FPS. Try renaming the exe.

Just Renaming Executable Prooves Nothing

[Mooncaller]

As a writer of low level instrumentation drivers, I have often played around with adaptive optimization. It realy is trivial for a driver to asay the optimization appropriate during an executables run. Basicaly application specific optimisation on-the -fly. The problem is that doing this slows the driver down more then the optimizations speed it up. But if the info is stored, the next time the executable is run, the driver just needs to engage the optimizations it had learnt previously. The second run will be considerably faster. This type of driver could very well have behavior some have reported for nVidias drivers. Did the tester run the renamed executable with as many succesive runs as the originaly named executable? Probably not, they did one run and happily concluded the nVidias cheating.

Re:Just Renaming Executable Prooves Nothing

[sixb0nes]

Did you happen to look at the screenshots that compared the two screens, side-by-side .. and the differences between them?

Re:Just Renaming Executable Prooves Nothing

[Oswald]

Well, I know you wouldn't post without first reading the article, so you must have missed the part at the top of the second page where it says: Tests were run three times, and the results were averaged.

A question: If the NVidia drivers were behaving as you describe, do you not think that this information would have been well-publicized, in order that people would understand why the damn thing ran so slow each time they did something which corrupted/deleted the drivers application database? It seems like the kind of thing you would want peopel to know.

Re:Just Renaming Executable Prooves Nothing

[Nom du Keyboard]

application specific optimisation on-the -fly.

Sounds like a JIT compiler applied to graphics, and somewhat akin -- in a very general way -- to Transmeta's approach to x86 of being slow the first time and faster later.

WHat if I renamed an EXE to 3dmark.exe

[Anonymous Coward]

would that increaseits performance? ehhe

Are you still running a 486 with VLB bus too?

[Anonymous Coward]

Are you still running a 486 with VLB bus too?

Cheating 101

[verloren]

I'm surprised this comes up so often. I would have thought that by now NVidia and all the other companies would have big notices on the wall of their dev offices:

"When cheating benchmarks, DO NOT use the name of the benchmark app as an indicator"

rendering accuracy benchmark

[Tommy_S]

Last week when this topic was on slashdot somebody posted that its possible to run these benchmarks rendering them through a software driver that conforms 100% to the DX9 spec. Slow as hell but you end up with EXACTLY what you should get. Maybe the 3DMark authors could render all this stuff in software so they'll know exactly what SHOULD be rendered by a video card and then have a new section in their benchmarks that compares what should have been rendered to what actually was. So, you run the benchmark and you end up with something like this: Nvidia FX: frames per second: 187, accuracy of rendered image: 62% ATI Radeon: frames per second: 177, accuracy of rendered image: 94%

Re:rendering accuracy benchmark

[Nom du Keyboard]

the 3DMark authors could render all this stuff in software so they'll know exactly what SHOULD be rendered by a video card and then have a new section in their benchmarks that compares what should have been rendered to what actually was.

EXCELLENT SUGGESTION. I'd mod you up if I only had points.

Re:So What? Who Cares?

[Trepalium]

Except it's still cheating. If the "optimization" was done in such a way that it automatically re-ordered shader programs to run faster on it's architecture, without caring which source or which pixel shader was being sent, it would be an optimization. Finger printing a pixel shader, and replacing it if it matches, regardless of if it's really the same end result or not, is still cheating.

Now, go and sell that nasty ATI or NVIDIA video card, and get an SIS based one. I'm sure THEY don't cheat on benchmarks (at least not video card benchmarks).

Re:So What? Who Cares?

[zbuffered]

The mass market will buy the box with the prettier logo, The box with the lower price, or the box that magazine xyz (that they read) says is best.

Most of the people I know don't do that -- they say, "Hey ZB, which graphics card should I get?" and I say, "ATI Radeon 9700" and they say, "OK" and that's that. I assume anyone in the market for gaming graphics cards knows someone they can ask for advice.

There is a simpler solution...

[Anonymous Coward]

Appropriately enough, the QOTD at the bottom of the page when I read this thread was:

The bogosity meter just pegged. ...and indeed it has! Goddamnit, this is not the American way! Both of the major video card contenders have been shown to cheat just to get a better score on benchmarks. They cheat and they lie and they do anything, anything, EXCEPT produce a better product.

This IS bogo(u?)s!

Its like a RACE

[Anonymous Coward]

In a race everyone is on the same track so you can judge who is the better driver and who has the better car, now if a driver uses a short cut to get to the finish faster, HE IS CHEATING, sure to the people at the finish line it looks like he is faster and better but in fact just cheated to get there....

SolitarySoviet

Polical Correctness

[oaf357]

So when students these days are caught cheating on tests can their excuse be, "Well, I was just optimizing my performance."?

Or better yet. If the kid renames his test is he not cheating?

Re:I will take a wait and see attitude. [OT]

[untaken_name]

Have you ever tried triggers under the XBox controller? You just hate to use DualShock2 controller ever again.

No, I probably should try the new controller, but all I did was pick up the original controller, then put it right back down. I couldn't stand it.

DualShock2 (and any controller that comes with any game console) is something that allows you to play all games but it's not perferct for anything.

I didn't say it was perfect for everything....I should have phrased it like this: Barring specialty (flight sticks, steering wheels, guns, dance pads, etc) controllers, the DualShock2 is the best controller out there for most games.

And all the first person shooters... anybody that claims that console controller is more suitable for that genre of games than keyboard and mouse should go online with some game that allows plaing against PC players and get their feet wet.

Yes, that's why I said they were the worst for FPSs.

One thing I especially hate with game console controllers is that those're designed to be used with thumbs only. Hello? See, I've four other fingers in both hands, why not use those? (Yep, some games support triggers but still using all 10 buttons in the DualShock controller is real pain in the ass.)

That's why most games let you configure the buttons. I know they don't *all* let you map *every* button, but that's the fault of the game designers, who make the control schemes less than intuitive and then don't let you change them. I have found that in almost every game, changing the default controls such that the most important commands are on the shoulders, instead of the pad, helps me immensely. This lends itself especially well to fighters like tekken and virtua fighter, as i have *no* problem hitting any combo of those four buttons. I map less used commands to the pad, because then I can use my thumb to hit them easily. This control scheme has serverd me well in every game that supports button mapping. (which is most of the better ones.)

The problem with all the consoles is lowest common dominator: if something doesn't come with every console it's supported by practically no game. How many games support, for example, more than 2 players on PS2 even though the console can support 8?

Most sports games do. However the reason more games don't support more than 2 players isn't really as much about what comes with the console. If you have 4 split screens, it can be very annoying to play. Games like Gauntlet let you have 4 people on the screen at the same time, but then you had to coordinate to move around in the world. Most sports games let you have 4 or 8 players, and they're probably the ones you most want to do that with. Fighting games that support more than 2 players are pretty rare anywhere, not just on consoles...driving games becaome unplayable with a 4 player splitscreen, unless you're rich. I personally would rather wait my turn than only get to see a 6" square on the screen. Also, there are a lot of 1 player games out there. This isn't because the console only has 2 controller ports. They're just single player games. There's never going to be a 'perfect' controller that is perfect for every type of game. Having said that, compared to every controller I've personally used heavily, which include the atari joystick, the intellivision pad thing that you hadda put overlays on, colecovision, commodore64 and 128, nes, super nes, sega master system, genesis, saturn, turbographics16, Playstation, N64, PS2, Gamecube, and many many 3rd party controllers for various of the above, as well as gamepads for the pc from a variety of companies, logitech, m$, belkin, various oem stuff...I've never owned a gameboy, don't own an xbox, and I never had an amiga...but other than that, if it's a console controller, I've probably worn at least one out.
Well, basically I've always loved video games of every kind, and played them whenever I could. I realize that having no life whatsoever has both pros and cons

NVIDIA's argument about optimizations is logical

[outer0rb]

Ok, most people don't understand that there is no one correct method for everything. NVIDIA has it's own rendering path for AF. They want 3dmark to take advantage of their optimization, so they detect when 3dmark is run and force it to take the faster rendering path. The net result is an undetectable quality difference (i can't tell the difference at all) -- and not necessarily quality degredation...just difference. It's not like when ATI made it's hacks so that quake 3 was visually horrible on ati cards. Honestly, all this hoopla is just that. 3dmark's use of generic rendering paths does not properly benchmark the abilities of the vid cards that have optimized paths. Obviously, game developers will take advantage of nvidia's rendering path for AF on nvidia cards and ati's rendering path on ati cards and default to the generic path for everything else. 3dmark is not a good method of benchmarking video cards. your favorite game is the best benchmark. obviously, there need to be some changes made (at 3dmark, nvidia, and ati) to address using vendor specific optimizations. and sensationalist journalists like this need to leave their biases out of their reports (or at least think critically about what they're reporting and realize that the larger issue is the innapropriateness of 3dmark as a proper benchmarking utility).

Running strings on driver

[HalfFlat]

Has anyone run strings (or its Windows equivalent) on the driver binary, and seen if any well known executable names come up?

If they do, it will give us a list of programs where Nvidia is doing application specific optimizations, or (as in the 3dmark case) are doing a dirty and substituting faster but wrong operations.

If they don't then it seems clear that they're deliberately disguising the fact, given that there's empirical proof they're doing it for 3dmark, and this would demonstrate that Nvidia are fully aware that what they're doing is underhand.

Re:Running strings on driver

[Anonymous Coward]

There is no equivalent for "strings" on Windows.

Re:and they all told me I was crazy..Suspend Them!

[Nom du Keyboard]

So NVIDIA is guilty of using corked drivers

So suspend them for 8 games, unless they are a big superstar and appeal.

And go back in the archives to see if they ever said, "You can test me for steriods any time -- except right now, of course."

Cork = Profit

[Nom du Keyboard]

1. Cork driver for performance.
2. If caught, claim it was only a practice driver sent in for benchmarking by mistake.
3. Profit!

Are you a stupid fucking shit?

[Anonymous Coward]

Are you a stupid fucking shit?

Re:So What? Who Cares?

[confused one]

Why is it sad? This is why:

The mass market will buy the box with the prettier logo

Most of the time, I don't care either; as long as it does what I want. I'm not out on a witch hunt... Besides, I was replying to Aren't we smart enough not to be pulled in my marketing hyperzor? and that's exactly what most people do, fall for the market hype.

Who and Why?

[Mystiq]

I'd like to know who decides on doing this or not. It's detrimental to no one but nVidia's own sales once they get caught and there's a whole lot of bad press that gets generated. I simply can't see a point other than the fact that they're behind ATI and desperate times call for desperate measures (sue me, I used an old saying).

I've never taken kindly to nVidia ever since the TNT2 and have always thought their method of business to be shady. They've always done nothing but try to increase 3D Mark scores and this is the height (bottom?) of their fight. And don't get me started on them releasing that new driver set so long ago to coincide with ATI's release of a new video card to overshadow it.

I don't like business who act shady. You can imagine my feelings towards Microsoft.

And yes, I'm still bitter over 3dfx. :)

Re:Who and Why?

[Mystiq]

Er, one more thing. Don't get me wrong, I was hopping mad when I found out about Quake/Quack but now nVidia has done that and more.

Optimizations that change your preferences behind your back aren't optimizations, they're deceptions. If I set anisotropic aliasing to 32x, I want 32x, not 0x. That's what Quake/Quack and 3DMark/3DMurk are and you can't argue that.

Oh and ever wonder why they specifically chose anisotropic aliasing and not, say, texture bias? I'm not saying this is it, but who remembers back when nVidia was bashing ATI for their anistropic filtering method? Are they trying to get them back or something? o.O

At the very least, as people are saying, ATI honed up to it right away. This is the least nVidia could have done and not label them as "bugs" or "application-specific optimizations." Even if this was done in a game (some people are calling 3DMurk a legit application-specific optimization), changing a user's options behind your back isn't an optimization.

How would you like if you disabled Java or Javascript in IE for all security clearances and when you go visit *.microsoft.com, it's enabled without you doing anything?

Re:So What? Who Cares?

[confused one]

It's good your friends trust you.

I suspect that your wrong though. Just because someone asks me "Which graphics card should I get?" doesn't make me an expert in graphics cards. I've a degree in physics, work as an electrical engineer and do instrumentation and controls programming. How does that make me an expert? It doesn't. (sometimes knowledge is knowing when you DON'T know something).

Oops, I was ranting...

The sad truth is that most people don't ask someone knowledgeable. They'll ask a salesman (who's looking for a commission). They'll ask the neighbor who owns a computer. They'll take what's in one particular magazine at face value, without checking it out themselves. (doing a simple google search...)

Sure, it's usually safe to pick some reputable brands "latest and greatest," which seems to be your advice (the Radeon 9700) But is that what they really want or need. I could buy a Porsche; but I drive an old Dodge truck because it gets me there (and the occasional rack full of electronic gear fits in it better).

Re:So What? Who Cares?

[R0]

Maybe that's one of the real reasons why nvidia refuse to help free drivers from being developed.

whatever

[kosamae]

Ummmmm, maybe its just me, but it was nearly impossible to see the difference from looking at ONE SPECIFIC "GARBLED" FRAME of this thing running. When ATI did their cheat, it was blatantly obvious just from comparing any two screenshots of gameplay. I don't know about anyone else here, but I'll sacrifice having every one of my frames being absolutely perfect (even after I run them through filters in gimp), for 10 extra FPS.

Re:and they all told me I was crazy..Suspend Them!

[Mysticalfruit]

so if an executable prefetches libraries, is that called "executable doping"?

I can see in now...
--headline--
Microsoft admits to executable doping, anonymous sources heard talking about how they "knew" their executables loaded too quickly. Microsoft spokesperson read a public apology to all executables that had ben resource starved because of the malaction of their executables.

No, don't mod parent up :-)

[Performer Guy]

The impact anisotropic settings are heavily dependent on the viewing angle of the texture. That particular screenshot does not appear to have particularly onerous viewing angles on any of the polygons. Given that the test setting is 8:1 anisotropic filtering you really require a view from the same demo where there's a steep slope on a few polygons to see the real visual impact. It doesn't look like any portion of that scene that's illuminated has more that 2:1 or maybe 3:1 anisotropic texture derivitive. I suspect what NVIDIA has done here is crank their aniso filtering down to maybe 4:1 in the driver, giving lower quality at steep viewing angles (blurrier textures on sloping surfaces), saving time through fewer texture filter taps and much better texture cache behaviour (due to lower resolution MIP LOD fetches).

Re:No, don't mod parent up :-)

[Lightman_73]

It's indeed possible. But as long as there is no other evidence, we can't really tell anything about it. And bashing nV without evidence of a significative worse quality seems a little wrong to me... ;)

Now, couldn't someone with the latest buildof 3DMark2003 post some more interesting and to the point image comparisons ? (I would, but, 1st, I don't have 3DM'03, and 2nd, I don't have a GF FX ;)).

Re:No, don't mod parent up :-)

[Performer Guy]

Dude, the evidence is there, and it's clear, I know a fair bit about anisotropic filtering. I also know that a conditional on an application name that mysteriously accelerates only for anisotropic filtering for one application is dickering with the filter taps and texture LODs and that is supported by the visual differences when viewed with a skilled eye. As an aside there are issues with temporal aliasing quality here too that you can exploit by tapping fewer samples and pushing the spacing and/or LOD that won't show up in a still. You can see the difference in places on the screenshot, it's just not as obvious as it would be if the reviewer had the sense to take a screenshot from a more illustrative position. The evidence is there, you can choose not to see it, or somehow reject it as not enough, but a sneaky anisotropic filtering lie where 8x is claimed but 8x isn't delivered is exactly the kind of dishonest shortcut that is subtle and difficult to detect. This is not just some ambiguous issue. When the dial says 8x aniso it isn't doing 8x aniso, it's a lie and it's it's a cheat.

We have 9 smoking guns with NVIDIA prints on them. This is THE most extensive set of cheats perpetrated in the history of graphics (and to heck with the joint press releases, I know what a cheat is) I've seen Futuremark's Q&A where they clearly say these "application specific optimizations" were unacceptable, even after the joint press release.

Re:So What? Who Cares?

[Glock27]

95% all games released do absolutely nothing that requires a GeForce 3 or better, and the simple reason is that the vast majority of PCs sold come with lesser cards.

Nope. Most of the big games contain multiple code paths, each optimized for a certain level of hardware. The game auto-detects the card's capability, then uses the correct path.

It is a pain, but it works...and the folks with fancy new cards get fancy eye candy. That makes for good reviews, which makes for good game sales. Most folk might not have bleeding edge hardware, but the reviewers do. ;-)

Real link

[neonfrog]

That was one of the most entertaining and interesting reads I've come across in a while.

Here's a clickable link:

Why Your Framerate Affects Jumping

Yeah, so?

[Dolemite_the_Wiz]

When is Nvida going to pull their collective heads out of their rectums and add methods for Gamers to increase their refresh rates and their frames per second on XP?

This is the biggest feature missing from their products and needs to be addressed!

Dolemite
_________________________

while I don't work at dell

[Anonymous Coward]

I think I should be fired if they were paying me $50,000 / year just to look at 3DMark scores and pick the highest one. Talk about being replaced by a very small shell script!

There's a lot of dancing behind the scenes of picking OEM components. 3DMark has respect, so you know people will take it into consideration. But there's also corporate relations, user feedback, demand, support history, etc. Why do you think there's no Dell computers with AMD even though AMD has won the speed crown several times from Intel? Because there's more than one benchmark invovled, that's why.

So?

[wuice]

Doesn't every video card company do this? I know ATI have been busted for doctoring the results before. What's the outrage? I bought an ATI card a while ago and I've had nothing but problems with it. I wish my 128 meg Radeon 8500 had been half as headache-free as my Geforce 3. To me, that's worth a lot more than 3dmark tests.

That's not pseudocode. BASIC is pseudocode.

[Anonymous Coward]

You typed C and is a valid representation of a possible solution because gen_random_filename() and benchmark_suite() are valid function calls that have not been revealed.

Obviously, you have a patent on gen_random_filename() and benchmark_suite() thus is why you didn't reveal to us their code.

True to what I typed in the subject line, BASIC is valid pseudocode because it lacks a consistent syntax in its many various implementations.

Return it and snipe yerself an S3 4MB from eBay

[Anonymous Coward]

Nethack doesn't need that fancy Direct3D or openGL for cool graphics. Check out NetHack text 3D

Re:So What? Who Cares?

[zbuffered]

I was generalizing. I own a 9500 'coz it's cheaper and does what I want. I'm a gamer. I play games. You, as a non-gamer, may not know what graphics card to get, but gamers will know. So ask a gamer. If you need to game. Otherwise get something cheap.

Apply DMCA!!!

[pkinetics]

Why don't the people who developed the benchmark software go after NVIDIA and threaten them with a DMCA lawsuit violation? Their going in and tweaking the parameters of the test... Isn't that a violation?

Amazon to apply for new patent ;-)

[pkinetics]

In the latest technological breakthrough, Jeff Bezos' Amazon.com has applied for patents involving but not limited to:

A) an algorithm to modify baseline testing parameters

B) an algorithm to modify video display results

Further refinements on the patents to follow.

Re:Image Quality (MOD PARENT UP)

[Performer Guy]

Anisotropic filtering is a big win when viewing images under certain conditions. The conditions are not met in the screenshot, that's all. When viewed at an oblique angle anisotropic filtering becomes expensive and benefits quality most, and this is where it makes a huge difference.

Re:NVIDIA's argument about optimizations is logica

[Performer Guy]

Bollocks, anisotropic filtering is a simple integer, even programmatically when it's not set in the driver. However you can override it in the driver, again with a simple integer that specifies state. There's no special rendering 'path' it's done in the low level texture filtering hardware. Fragment shaders request the texture fetch and may have a 'path' but the actual aniso filtering is a low level pipelined hardware operation where the application specifies the state from a fixed function set of filter parameters. What NVIDIA did here was to ignore all those numbers, and simply cheat by reducing the anisotropic filter taps, basically they said they were doing 8X aniso, but they lied and did less, or pushed out the MIP LOD to help cache behaviour, either one is a flagrant cheat. There is no special rendering path, it's one integer specified for the texture filter. The screenshot does not show anisotropic filtering in action (actually texturing at an oblique angle), if it did and the author has simply picked a better view from the demo the quality difference would be apparent.

Re:So What? Who Cares?

[Sj0]

Why would they? If the card does what they want it to do, why bother searching the entire market for another alternative -- especially for a 5FPS difference in Game X(and a 5 FPS loss in Game Y)?

Re:So What? Who Cares?

[saden1]

Magazine? Who buys those anymore? Amazon.com and Google all the way baby!

Re:So What? Who Cares?

[Shardis]

Yeesh, and you my friend are one of the reasons why companies put out substandard products and try to hide that fact.

They're going for the people that are willing to pay out the big bucks for these features on their graphics cards to get the latest and greatest when it possibly turns out that all those shiny high benchmarks they were looking for were artificially inflated...

Someone smell something that's reminiscent of a bait and switch? That's the only thing that comes to my mind. That's if all of this is actually true, I haven't independantly verified anything ofc...

(offtopic)

[Shardis]

Rofl, I usually hate it when people quibble over spelling in forums and etc, but your "eke" comment actually made me laugh out loud. :)

Not a bad idea about the lawsuit either. If this type of thing gets to be pervasive (as it's seeming to) something has to break...

Re:So What? Who Cares?

[Anonymous Coward]

You aren't listening are you.

When you compile with +O2 you get the same result as if you did +O1, but the output is faster.

If you use an optimisation that swaps the 80-bit IEEE numbers for a fast MMX emulation of floating point, you DONT get the same answer (the precision is gone).

OK *IF* the output only required thelower resolution, but that should be disclosed.

ATI did remove their optimisation, maybe becayse they didn't know if it would help other games.