3D11 secrets?

[G.Foyle]

Do you see www.xtremeimagequality.com at the top of the page?

[Poci]

Do you see www.xtremeimagequality.com at the top of the page?

Doh I get it, but are you getting my point? If someone is looking JUST for the sake of comparison at the performance of a card, there should be a standard for comparison. Yes I get it, get max benchie score through driver tweaks. I simply think thats an easy way out and then others will make a million posts wondering why the card they have thats the same as yours isnt performing as good.

[Gautam]

Please play nice guys.

[K404]

Doh I get it, but are you getting my point? If someone is looking JUST for the sake of comparison at the performance of a card, there should be a standard for comparison. Yes I get it, get max benchie score through driver tweaks. I simply think thats an easy way out and then others will make a million posts wondering why the card they have thats the same as yours isnt performing as good.

Well.... all our scores are standardised within the same set of rules.

I suppose the most important question I could ask you is.... how can driver settings be enforced across thousands of results by hundreds of benchers who have never met each other? ?

[ZenEffect]

Poci: what driver tweaks resulted in a noticable gain in points for you? AFIK 3DMark11 does not react to any driver tweaks. Just re-tested using 580 and saw no gains that I can attribute to tweaking with inspector. 40 points was what I saw but that was due to running the bench repeatedly and the inconsistancy of the benchmark.

[chew*]

My uber tweak, = swap out all my monitors for the one that scores the highest........yes I am killing my graphical experience by using crappy but high scoring monitors. Fear my 90LB ancient CRT people.

[Poci]

Well.... all our scores are standardised within the same set of rules.

I suppose the most important question I could ask you is.... how can driver settings be enforced across thousands of results by hundreds of benchers who have never met each other? ?

Well, In the perfect benchie comparison utopia, I think we should all click the reset to factory defaults on ccc or nvidia control panel and then compare. OCage on the other hand is a differnt story, I get a bigger satisfaction from seeing my scores bump up from that.

Poci: what driver tweaks resulted in a noticable gain in points for you? AFIK 3DMark11 does not react to any driver tweaks. Just re-tested using 580 and saw no gains that I can attribute to tweaking with inspector. 40 points was what I saw but that was due to running the bench repeatedly and the inconsistancy of the benchmark.

Actually I havent tweaked a driver for ages for benching, last time I tryed it just to see the difference was with prolly 3dmark 01 or 03....... And I did notice an improvment. But then I figured hmm, I could lower rez to 480 and min every posible quality setting to performance, but then I wouldnt be stressing my video, I would be stressing CPU.......

[K404]

Well, you're asking everyone to give up free performance from a setting who's use can never be proven or disproven- you're asking for 100% global honesty and when there are rewards for the higher score, that won't happen.

Also, even if we overcome THAT, all nVidia and ATI have to do is adjust their driver settings and the default options.... and I hope you know as well as everyone else, they are more than happy to do that to improve review performance.


The idea is noble and transparent.... and totally unobtainable.

[Poci]

nVidia and ATI have to do is adjust their driver settings and the default options.... and I hope you know as well as everyone else, they are more than happy to do that to improve review performance.

I know very well both nvidia and ati have been doing that for ages. They do minor tweaks to the factory defaults usually to the textures setting for those few extra points.

[T_M]

You are spewing a load of crap, sorry to say it (actually..... no im not).
But then again, there will be a mod coming along soon to say I shouldnt say that.
In the meantime i will point out a significant flaw in your argument - comparison of hardware performance across different driver revisions.

I also note that your first ever post was in this thread - after 2 years.
And your first post would be considered by many a trolling post.
Maybe you can share what made you suddenly decide to point out after 2 years that driver adjustment is bad?

[Poci]

You are spewing a load of crap, sorry to say it (actually..... no im not).
But then again, there will be a mod coming along soon to say I shouldnt say that.
In the meantime i will point out a significant flaw in your argument - comparison of hardware performance across different driver revisions.

I also note that your first ever post was in this thread - after 2 years.
And your first post would be considered by many a trolling post.
Maybe you can share what made you suddenly decide to point out after 2 years that driver adjustment is bad?

1st of all you do realize they just opened registration considering the 2 years.... it means I couldnt post.

As far as me spewing crap, I simply said that I know its true both ati and nvidia do tricks with their drivers for the benchie scores. How is that spewing crap? And I certainly can judge you should also know this already. I dont know what you consider trolling, but in my book its responses like what your giving me. My first post was in no way trolling and simply asking you all, what do you think of equal comparison playing fields. In any case, im looking at your sig right now, you should know what im talking about when it comes to what both nvidia and ati do to drivers to affect benchmark scores.

And I never said driver adjustment was bad, I said you could never acheive fair comparisons when one guy has his card tweaked one way and another guy with same card has it tweaked differently. What? you dont think I adjust my own drivers for gaming? I been doing that with rivatuner before nvidia control panel even gave us all the goodie options..... but benching is a differnt story...
In the end responses like yours are only showing a y attitude to newcomers, and whats even funnier is that you somehow assume Im a n00b. So then when you first registered here were you a n00b too? Gimme a break. Dont be so ignorant.

[Poci]

Well.... all our scores are standardised within the same set of rules.

I suppose the most important question I could ask you is.... how can driver settings be enforced across thousands of results by hundreds of benchers who have never met each other? ?

What specific rules do you guys use? And as for your question, how can driver settings be enforced? Simple, convince people to click defaults and then bench.

[K404]

http://hwbot.org/article/general_rules



Can't convince people to do something, when no-one will ever know they didn't do it.... sometimes, a benchmark ranking has too much hanging on it for the honourable thing to be at the forefront of peoples minds.

"Is this against the rules- yes, or no?" is sometimes the only question.... and the rules are mostly based on what can be *enforced*



This is not my personal opinion towards benchmarking.

But... I set GPU driver settings to highest performance, AA, AF off. I usually play FPS and... I don't stop to look at walls and marvel at how smooth the surfaces are.......

[[XC] gomeler]

What specific rules do you guys use? And as for your question, how can driver settings be enforced? Simple, convince people to click defaults and then bench.

No mipmap is really it. If you can adjust it in the driver panel offered by the manufacturers then it is generally considered a legit adjustment. If you want hardware comparisons with a gaming perspective then perhaps you should explore the Internet for websites that provide stock reviews. I hear there are many fine establishments out there.

Here we couldn't generally give a flying for image quality. I mean seriously.. in some benchmarks we're average thousands of FPS and rarely watching the screen. I just ran 3D11 and actually watched the benchmark for the first time this weekend. Looks like crap on default settings and I'd much rather be watching my thermometers/multimeters and the FPS indicator.

[ZenEffect]

I just ran 3D11 and actually watched the benchmark for the first time this weekend. Looks like crap on default settings and I'd much rather be watching my thermometers/multimeters and the FPS indicator.

wait, there is more than meters and fps counter?

poci: benching isnt always about pure hardware clocks, there is software, and a person involved as well all to total to a score. the fact that some people can run a benchmark faster due to their knowledge of the bench, operating system, drivers and hardware is what is fun about it. as much as id hate to say this, think of nascar. for the most part, all the cars are identical in spec, but the engineering team, pit crew, pit manager, driver etc all make the car faster to determine a winner. same goes with benchmarking. if it were left completely up to the hardware, then the richest people who overclock would dominate as it would become nothing more than a game of binning the best parts rather than tuning the system (its already more than half way there tbh... thanks intel )

[Poci]

I don't stop to look at walls and marvel at how smooth the surfaces are.......

Dude, I agree with you all that for benching there is no need for image quality. But in a gaming session itself Im not looking at walls, Im looking at bushes and mountain tops at a distance, and the higher the image quality the better chance my eye will have of seeing that sniper aiming at me.......

No mipmap is really it. If you can adjust it in the driver panel offered by the manufacturers then it is generally considered a legit adjustment. If you want hardware comparisons with a gaming perspective then perhaps you should explore the Internet for websites that provide stock reviews. I hear there are many fine establishments out there.

Here we couldn't generally give a flying for image quality. I mean seriously.. in some benchmarks we're average thousands of FPS and rarely watching the screen. I just ran 3D11 and actually watched the benchmark for the first time this weekend. Looks like crap on default settings and I'd much rather be watching my thermometers/multimeters and the FPS indicator.

Yes I know, and I have been follow lord K1ngp1n for years and I know this is his "home"

wait, there is more than meters and fps counter?

poci: benching isnt always about pure hardware clocks, there is software, and a person involved as well all to total to a score. the fact that some people can run a benchmark faster due to their knowledge of the bench, operating system, drivers and hardware is what is fun about it. as much as id hate to say this, think of nascar. for the most part, all the cars are identical in spec, but the engineering team, pit crew, pit manager, driver etc all make the car faster to determine a winner. same goes with benchmarking. if it were left completely up to the hardware, then the richest people who overclock would dominate as it would become nothing more than a game of binning the best parts rather than tuning the system (its already more than half way there tbh... thanks intel )

Correct, but some of us would rather rely on fate for a lucky chip, plus ordering from the egg or an online store for example you get no choice of what batch your getting. Unless we go to a retail store and happen to allow the employees to let us inspect all the chips in stock at store to see which one suits our needs.

[chew*]

Dude, I agree with you all that for benching there is no need for image quality. But in a gaming session itself Im not looking at walls, Im looking at bushes and mountain tops at a distance, and the higher the image quality the better chance my eye will have of seeing that sniper aiming at me.......

This could not be more wrong.

Single player eye candy up, I like to see my wall textures.

Anyone who seriously plays multiplayer and or games competitively especially clans turns all the crap off for many reasons. Higher FPS. less trees which means you can see snipers way better than full eyecandy.

Bottom line professional gamers shut all the eye candy off, it's less distracting that way, some even go as far as to modify the .cfg to the min allowed without triggering anticheat, PB etc etc.

I am running 2 6990's and still run crap settings for multiplayer in BC2, I can see further I can see through trees and or the tree line is far more recessed exposing people who think they are hidden..........

[SoF]

I am running 2 6990's and still run crap settings for multiplayer in BC2, I can see further I can see through trees and or the tree line is far more recessed exposing people who think they are hidden..........

and that on your uber-CRT - that is close to cheating I would say - you are wasting your oc-skillz in gayming...bad chewie!



j/k as usual

[ZenEffect]

Correct, but some of us would rather rely on fate for a lucky chip, plus ordering from the egg or an online store for example you get no choice of what batch your getting. Unless we go to a retail store and happen to allow the employees to let us inspect all the chips in stock at store to see which one suits our needs.

Getting a lucky chip requires alot of "fate." Thats the reason why we tune, for example there are only a handful of 5.8+ 2600k, and even less 5.9. as far as 6.0 goes, you can count them on one hand. Alot of times you have to play the hand you were delt. I havent been able to pick a batch since the Q6600 days, and even then, my options were extremely limited. In a way we all rely on fate, but we also rely on ourselves to keep things competitive.

[Janus67]

I've never had trouble picking a batch at Microcenter (at least the one near me in Columbus)

On the topic of driver changes, I mean shouldn't/wouldn't you also have to ban stripped OSes, etc as well? Specific service packs, builds, etc? It's just not possible. And without a video every time of the person hitting default settings then benching you are completely on the honor code to not go back and change to defaults before screenshotting. The tweaking is 50% of the fun imo - it is boring to just sit there and go 'oh, that's as far as it goes, OKAY -- WTS GTX 580'