NEW - 3DMark 11 Compilation

[yogurt_21]

this is why benching against yourself is more fun. Seeing if youre uber lucky "I can't believe it didn't crash during that run" was really all your rig had. To me if the clock isn't 100% stable but the bench completes, fair game. If there's a setting tweak that gives you an edge , fair game. (ala nvidia physx in vantage). If you bench it outside in negative 20 degrees, fair game. If you're using "pay to win" hardware, fair game. If you can pump throw mod after mod after mod at it, fair game. If it only works once, fair game.

and etc.


It's benching, not reviewing. Think of it like a quarter mile run. I don't care if you put the pistons through the block and the driveshaft snapped at the finish line. If you crossed, it counts.

 

[EarthDog]

Not complaining, nothing against the rules, sure, but at least mention you made the image quility into raw sh*t so people won't think that this is the performance under default settings.

LOL, you want "raw shit" as you so eloquently put? Try changing LOD settings for the older benchmarks. Lowering everything that we are, like tess and IQ, you still know EXACTLY what the scene is rendering. You mess with LOD, and you cant tell anything in the image.

Nice score Dom! thanks for rising up to the challenge

Although, IMHO I think turning tesselation off sort of defeats the purpose of running 3DMark 11 as it specifically mentions tesselation as a feature right before each test, so basically what you're doing is running it in DX10 mode perhaps we should open a different ladder for benchies run without tesselation, like we did for the ppl who use vantage with PhysX enabled

This is just a suggestion, not accusing anybody of cheating, just my $0.02, not looking to start an argument

I wonder how high my cards would score if Nvidia gave me the option of turning tesselation off, nothing wrong with some healthy competition, props to AMD for giving you guys that option, cheers

At Hwbot, where benchmarkers go to party and live by official rules, it is a point of contention as well. The problem they were having is that its an option in the control panel which they deemed legal.

The only people that have the option to run without tesselation are AMD users. Currently, easily, like what people here would know, Nvidia does not have the ability to disable Tesselation via software.

 

[cadaveca]

It's benching, not reviewing. Think of it like a quarter mile run. I don't care if you put the pistons through the block and the driveshaft snapped at the finish line. If you crossed, it counts.

That's exactly my point. My only request is that guys list the tweaks, and Dumo's doing just that. Most of them, like Dumo's turning off tesselation, are things I know about, and expect. Everyone should be using it; it gives me about 1500 points in this bench, with HD6950 lulz. I won't turn off tesselation for gaming.

DJ-Electric and I, being reviewers, use this bench for other things. I mean really, as a fair compare, you'd need to use the same driver and stuff. That's far too much work.

 

[EarthDog]

Im a reviewer and use this bench for other things as well. I also happen to be balls deep in benchmarking at hwbot. best of both worlds.

 

[erocker]

I have a degree from a University where reviewing and critiquing was a part of my syllabus. I own and operate a successful. business where I deal with specifications and tolerances constantly. I thought this was a share something about yourself time but it has absolutely nothing to do with this thread.

 

[dj-electric]

Can we just drop it? this argument seems really unnecessary...
let the guy post his results and get 5000K on pacman-quality image, i stopped caring by now.

 

[EarthDog]

It wasnt necessary to bring it up in the first place Dj...but you broke that up with your post about it in the first place. hahaha!

But before you continue to go off on 'pac-man' quality image, try disabling yourself and see the difference and lowering IQ. Its really not that much. IN heaven the biggest difference in IQ with tesselation gone is the size of the rocks in the first scene. Otherwise, good luck at noticing a difference (obviously there are some, but you have demonstrated dramatic license, so am I).

but that's how it works when people keep secrets.

Yeah like the disable tesselation thing is hidden in some super secret menu? or the "quality" vs "performance" isnt blatently obvious in CCC/NvCP either.

Anyhoo, back to the regularly scheduled programming. Reviewers go review since you dont know benchmarking well. Leave the benching to people that do it, sort of well.

 

[15th Warlock]

To be honest with you I didn't even know AMD had that advantage until Dom brought it up a few pages back, we can always do two ladders like we did for Vantage and PhysX, I think that would be the most fair thing to do, if ppl are getting over 1500 more points just by disabling tesselation like cadaveca says, then that makes a big difference in your results.

The way I've always benchmarked is by resetting all driver features to default before running any particular benchmark to make it a level field, but even then there are specific IQ settings that are enabled or disabled by default by either Nvidia and AMD, so I think we are probably never going to agree on this.

If AMD gives this option to their users, then the more power to them, I just agree with cadaveca about ppl reporting such "tweaks" like Dom did, we did the same for Vantage and PhysX and IBT and AVX, so why not do it with '11, that would put an end to this argument

 

[EarthDog]

Thats the problem with threads like these...the point is to get the highest score but those that are not in the know, start them expecting something different. Clearly, to me, the point was NEVER to measure apples to apples. As there are system differences from ram speed (which matters on some benchmarks... OH DID I GIVE OUT ANOTHER SECRET?!!!!), CPU speed, drivers, OS, etc. So if those who start threads like this, or even THOUGHT threads like this could compare ANYTHING outside of the who can get the highest score, is crazy talk. If that was the goal, it should have been stated with rules defined from the get go. Otherwise, as it should be, its a free for all to get the highest scores you can in any way possible. Be it extreme cooling (wonder why that wasnt brought up?) Different drivers, different settings in the drivers, etc. I mean who goes to the drag strip to race against ONLY stock camaros oh yeah, thats called bracket racing (you get the point - bracket racing = similar times, not cars).

My advice is to leave this thread as is. And if you want something to compare settings to, start another thread. GL.

 

[15th Warlock]

Thats the problem with threads like these...the point is to get the highest score but those that are not in the know, start them expecting something different. Clearly, to me, the point was NEVER to measure apples to apples. As there are system differences from ram speed (which matters on some benchmarks... OH DID I GIVE OUT ANOTHER SECRET?!!!!), CPU speed, drivers, OS, etc. So if those who start threads like this, or even THOUGHT threads like this could compare ANYTHING outside of the who can get the highest score, is crazy talk. If that was the goal, it should have been stated with rules defined from the get go. Otherwise, as it should be, its a free for all to get the highest scores you can in any way possible. Be it extreme cooling (wonder why that wasnt brought up?) Different drivers, different settings in the drivers, etc. I mean who goes to the drag strip to race against ONLY stock camaros oh yeah, thats called bracket racing (you get the point).

My advice is to leave this thread as is. And if you want something to compare settings to, start another thread. GL.

Sure, it's not my thread, and there's no rules against these "tweaks" so, whatever, I'll just keep posting my results in it besides, I run these benches for the thrill of seeing my system pushed to the max without crashing, like yogurt says comparing my results to others is just icing on the cake like I said before, nothing bad with some healthy competition.

I was just making a suggestion to end the argument, I'm not gonna start a new bench thread just because of that, so whatever floats your boat I guess

Peace

 

[DOM]

Wait till I put them on my 5.8 2600k boohaha

 

[SonDa5]

At the most competitive level it does seem like an unfair advantage if AMD has an option that will increase score of this benchmark.


Since we all know about this AMD trick to increase score I think it would help if AMD card owners share if they disable tesselation in their scores.

 

[t_ski]

At the most competitive level it does seem like an unfair advantage if AMD has an option that will increase score of this benchmark.

I would say that is not true, as Nvidia cards have for a long time had a higher score advantage due to the physx processing onboard.

 

[SonDa5]

I would say that is not true, as Nvidia cards have for a long time had a higher score advantage due to the physx processing onboard.

Look what happened with Vantage. The bench marking community came to the conclusion that it was fair to rewrite vantage so as to make it a fair competition.

 

[DOM]

If you don't like it why don't you ppl make your own ?

I'm sure its already a pita to get this one updated why make him work harder ??

 

[brandonwh64]

 

[EarthDog]

At the most competitive level it does seem like an unfair advantage if AMD has an option that will increase score of this benchmark.


Since we all know about this AMD trick to increase score I think it would help if AMD card owners share if they disable tesselation in their scores.

Thats funny, myself and DOM compete at the 'highest level' in that we both have used LN2 and the like and participate at Hwbot.org. Its legal there. Its not liked per say, but seeing as how its a CCC function, its legal there.

Leave this thread as is.

 

[Outback Bronze]

LOL. When is debating ever going to end. It doesnt matter what subject it is, theres always a debate. Whos right and whos wrong? Does it really matter? Let the debate continue...

 

[15th Warlock]

There's been over three pages of arguing over the validity of the results, this thread was supposed to be here for us to post our results and nothing else, got something else to say post it in another thread, let's move on people.

This is a quote from the very first post in this thread to remind you guys why we are here:

This thread is for posting your scores, not complaining about the benchmark or its methods.

 

[EarthDog]

I blame Dj for starting it...He opened the flood gates.

 

[DOM]

well the new 2600K is hooked up to the SS, but dont think the 80f+ in the garage is going to let me oc the cards to high

http://valid.canardpc.com/show_oc.php?id=2327512

 

[EarthDog]

Texas summers... Its still 65F in the (finished) basement.

Lets see what 55x and 7870 will do sometime.

 

[ZenZimZaliben]

DJ-Electric and I, being reviewers, use this bench for other things. I mean really, as a fair compare, you'd need to use the same driver and stuff. That's far too much work.

You're not a reviewer, so I do not expect you to take it seriously, but when results are put into question, I have interest, period.

Since I am now doing reviews, and want to relate performance as the end user will get when gaming or other such "normal" things, I've stopped "extreme benchmarking" for scores, stopped posting in threads like this, and disabled my score reporting in HWBOT.

So hey, Cadaveca...Are you a reviewer?

Hey look everybody... Cadaveca is a reviewer! He reviews stuff. On the interwebs. Reviewing things is what he does.

LOL. Just B/S'ing. Funny stuff.

Also, I am re-benching my rig with all these tweeks I really thought we were supposed to run default.

 

[DOM]

you are running it a default, these are driver settings nothing special...

and i failed yesterday it was to hot to run the 2 cards on stock cooling most i got was 1125 17k

might bring in the lil bench table into the livingroom

 

[cadaveca]

So hey, Cadaveca...Are you a reviewer?

Hey look everybody... Cadaveca is a reviewer! He reviews stuff. On the interwebs. Reviewing things is what he does.

LOL. Just B/S'ing. Funny stuff.

Also, I am re-benching my rig with all these tweeks I really thought we were supposed to run default.

You jelly?

At least I recognized what DOM was doing from the get go. Since, because I review hardware, I run this bench nearly 100 times a week, if not more, and I know it pretty well.

The whole issue over tweaking, even on HWBOT, is that tweaks need to be shared, so they can be validated as genuine. This is the pursuit of points, after all, not image quality. Turning off tesselation should be a given. I did for my own score in here with my 5870.