[H] Battlefield 3 Open Beta Performance and Image Quality

[Grooveriding]

[H]ard has put up their BF3 performance benchies

http://hardocp.com/article/2011/10/03/battlefield_3_open_beta_performance_image_quality/

He also concludes that performance is not consistent enough to get accurate benches

'Performance in the Battlefield 3 public beta inconsistent. The multiplayer nature of the beta made it impractical to objectively compare performance between video cards, but we did see some relatively repetitive behavior among video card brands. In general, we had better performance using video cards equipped with AMD GPUs than we did with NVIDIA GPUs. We downloaded the newest beta drivers from both GPU makers specifically released for this game. We tested each video card thoroughly, and the results with AMD video cards were faster, but not by much in this open beta.

On the official BF3 Blog, EA/DICE staffer Joe Ellis pointed out that "The beta won’t include all the graphics features which will make it in the final version". When the final retail build is released, everything we've seen so far could be turned around completely. But for now, single-GPU gamers with AMD Radeon HD 6000 series video cards will probably see slightly better performance than gamers with NVIDIA GeForce GTX 500 series counterparts.'

 

[RobertR1]

They should have disabled HBAO/SSAO on the 580. It helps out framerates a lot at 2560x1600.

 

[blackened23]

So what happened to all the nvda users who claimed the GTX 570 was 2x faster than the 6970 in bf3?

Hmm! Another mystery solved.

 

[houe]

Are the nvidia cards running out of vram? These results seem to be reverse of what most others are saying...

 

[skipsneeky2]

I believe the lack of vram compared to the 6950 and 6970 gives the amd cards the upper edge....

Article failed to mention how much vram was used if not much i guess nvidia has some catching up to do driver wise for this game.

Someone somewhere needs to make a article or thread posting vram usage at 1080p using low/medium/high/ultra aa and no aa included.

Doubt many gamer's will game at 1680x1050 and if so the choices are more obvious there .

Edit:Article needs to be revised with 3gb gtx580/2.5gb gtx570 cards for a more fair benchmark....

 

[blackened23]

Hmm, yeah. Maybe they should put some 1080p results up as well.

 

[blackened23]

They actually have 1080p results. A note on this, I play at 1080p and my VRAM usage according to MSI afterburner has never exceeded the 1-1.1gb VRAM range. Which is pretty unbelievable to me, i'd expect a game using this much VRAM to look *much* better.

Its totally plausible that at 2560x1440, that the nvidia cards would run into VRAM limitations. I do not believe thats the case at 1080p, at least in my experience. Should be a fair fight:

 

[ViRGE]

Humm, it is possible to publish articles just to troll readers?

The game is a beta - a DICE beta at that - it's not optimized, it's not even shipping with the highest quality textures or all of its DX11 features enabled. Basically nothing in this article will be relevant for the shipping version of BF3.

 

[Grooveriding]

I believe the lack of vram compared to the 6950 and 6970 gives the amd cards the upper edge....

Article failed to mention how much vram was used if not much i guess nvidia has some catching up to do driver wise for this game.

Someone somewhere needs to make a article or thread posting vram usage at 1080p using low/medium/high/ultra aa and no aa included.

Doubt many gamer's will game at 1680x1050 and if so the choices are more obvious there .

Edit:Article needs to be revised with 3gb gtx580/2.5gb gtx570 cards for a more fair benchmark....

I am waiting on seeing an article addressing the VRAM usage in this game. This is the heaviest usage of VRAM I have seen in any game. VR-zone made note of the heavy memory usage - http://vr-zone.com/articles/battlef...reenshots-and-hd-video-walkthrough/13625.html

Still want to see something comprehensive though. There are so many settings to play with, it would be a chore, don't want to test it myself :sneaky: I have noticed the deferred MSAA, ultra textures and ambient occlusion (HBAO/SBAO) settings are the biggest memory hogs.

 

[happy medium]

Humm, it is possible to publish articles just to troll readers?

The game is a beta - a DICE beta at that - it's not optimized, it's not even shipping with the highest quality textures or all of its DX11 features enabled. Basically nothing in this article will be relevant for the shipping version of BF3.

"stands up and cheers this post" :thumbsup:

Whats the excuse now why Hardocp's benchmarks once again are the total opposite of every other site out there. The old excuse was canned benchmarks vs real world gameplay.

Last I checked there is no canned bench for BF3.

I said it once and I'll say it again, Hardocp's biased, controversial, and more often then not, crap benchmarks are not worth reading, as is this thread.:thumbsdown:

 

[RussianSensation]

I am now officially convinced that HardOCP is biased in favour of AMD. HD6950 is 10% faster than a GTX570 at 1920x1200? Ya, sure...In every new game, his website is the only one that continues to show AMD in a much better light than anywhere else. Coincidence? Same website that claims MLAA has awesome image quality when it results in texture blur in half the games.

http://www.pcper.com/reviews/Graphi...esting-and-Image-Quality-Evaluation-Day-1/Ind

http://www.pcgameshardware.com/aid,...cards-SLI-und-Crossfire-benchmarked/Practice/

http://gamegpu.ru/Action-/-FPS-/-TPS/Battlefield-3-Beta-test-GPU.html

3 reputable websites vs. Kyle's world.

I am going to wait until a real website does BF3 benchmarks once:

1) The final game is out
2) Both camps have had a chance to optimize drivers
3) A patch or 2 is released for the game

 

[superccs]

This game sure kicks my OC'd 460 in the balls. 100% gpu and memory usage nearly all the time.

 

[toyota]

hardocp sure does get some odd results. why is there never another site that can get the same outcome that they do???

 

[blackened23]

I am now officially convinced that HardOCP is biased in favour of AMD. HD6950 is 10% faster than a GTX570 at 1920x1200? Ya sure.

http://www.pcper.com/reviews/Graphi...esting-and-Image-Quality-Evaluation-Day-1/Ind

Maybe in Kyle's world.

I am going to wait until a real website does BF3 benchmarks.

Wow. Unbelievable. :thumbsdown:

Whats funny is that the benchmarks from the alpha were similar. Anyway, It probably has something to do with HardOCP using beta drivers for both boards, something that previous benchmarks did not do. The beta drivers boost AMD performance fairly substantially - numerous posts on the BF3 boards confirm this.

 

[RussianSensation]

Wow. Unbelievable. :thumbsdown:

Whats funny is that the benchmarks from the alpha were similar - 6970 faster than 570. I guess the only results out of line were the ones you posted, russian. But since it favors AMD , Hardocp is lying. Interesting. Carry on.

It probably has something to do with HardOCP using beta drivers for both boards, something that previous benchmarks did not do. The beta drivers boost AMD performance fairly substantially. But I won't let this ruin your conspiracy theory.

Did you check the reviews?

PC Perspective: "We used the latest drivers from both NVIDIA and AMD that were released specifically yesterday for Battlefield 3: 285.38 for NVIDIA and 11.10 Preview for AMD."

PCgameshardware: "Since we're using the latest drivers from AMD and Nvidia, the Catalyst 11.10 Preview and Geforce 285.38, these results are state-of-the-art."

Gamegpu.ru: "NVIDIA GeForce / ION Driver Release 285.38 / AMD Catalyst 11.10"

Can you please show me other reviews/benchmarks where an HD6950 outperforms a GTX570 by 10% at 1920x1080 or 1920x1200.

Let's see now, he uses 4x MSAA + post processing AA? Who does that? Why would you apply both?

Also, his magical 6950 is 10% faster than a GTX570 while everyone else is showing it far behind? Notice in every review Kyle does, AMD cards do ridiculously well, despite everyone else showing otherwise.

My point is this isn't the only game. The same story continues 'over there'. HD6970 always performs as good or better than a GTX580 in his testing. And when it's slightly slower, he throws the price/performance card. Basically, AMD cards always win according to him.

So you have 3 reputable websites showing GTX570 easily beating an HD6950 2GB and HardOCP showing it 10% faster......

 

[happy medium]

I am waiting on seeing an article addressing the VRAM usage in this game

I did my own looking into this and at 1080p ultra , I dont a see significant drop in my fps. Don't take my word for it

 

[blackened23]

http://www.hitechlegion.com/reviews/gaming-software/13498?start=2

All cards using latest beta drivers supporting BF3:

I think its fairly plausible based on this, that the 6950 and 570 are pretty close.

 

[Grooveriding]

I am now officially convinced that HardOCP is biased in favour of AMD. HD6950 is 10% faster than a GTX570 at 1920x1200? Ya, sure...In every new game, his website is the only one that continues to show AMD in a much better light than anywhere else. Coincidence? Same website that claims MLAA has awesome image quality when it results in texture blur in half the games.

http://www.pcper.com/reviews/Graphi...esting-and-Image-Quality-Evaluation-Day-1/Ind

http://www.pcgameshardware.com/aid,...cards-SLI-und-Crossfire-benchmarked/Practice/

http://gamegpu.ru/Action-/-FPS-/-TPS/Battlefield-3-Beta-test-GPU.html

3 reputable websites vs. Kyle's world.

I am going to wait until a real website does BF3 benchmarks once:

1) The final game is out
2) Both camps have had a chance to optimize drivers
3) A patch or 2 is released for the game

Well Russian he does qualify that he feels the beta is useless for accurate benchmarks in his conclusion.

How can anyone expect consistency across different benchmarks from different sites. There is no built-in benchmark along with there being no single-player. All anyone can bench is multiplayer gameplay which will be completely different every time. As well, the one map that can be benched is much lighter on video cards than the big open Battlefield-esque maps are/will be.

I think the review is excellent for highlighting that there is no solid conclusion on where all the different cards end up, as right now there are circumstances where AMD is faster and ones where nvidia is faster. Until there can be a consistently identical benchmark performed, it's up in the air.

I can't speak to two of those sites, but pcper is pretty much a worthless site being blatantly biased towards nvidia for a long time now.

 

[happy medium]

Wow. Unbelievable. :thumbsdown:

Whats funny is that the benchmarks from the alpha were similar. I guess the only results out of line were the ones you posted, russian. Anyway, It probably has something to do with HardOCP using beta drivers for both boards, something that previous benchmarks did not do. The beta drivers boost AMD performance fairly substantially - numerous posts on the BF3 boards confirm this. But I won't let this ruin your conspiracy theory, obviously anything not favoring nvidia is biased.

Stop making excuses for this worthless website, its getting ridiculous.

I stood up weeks ago and shouted at the top of my lungs this site sucks and is biased and no one believed me.

I think others are starting to finally see the light, about freaking time!:thumbsup:

 

[blackened23]

See above results from HTL, Happymedium. The new AMD catalyst release boosted BF3 performance a ton....

Anyway, having played it on both sets of hardware (580 - 6970), I think they both run it fine. I don't understand how its hard to believe that the 6950 is competetive with 570? Because in most games they offer similar performance. The exceptions are crysis 2 and a couple of other TWIMTBP titles - which isn't hard to understand when some of those games force AMD hardware to do post processed effects.

 

[3DVagabond]

Humm, it is possible to publish articles just to troll readers?

The game is a beta - a DICE beta at that - it's not optimized, it's not even shipping with the highest quality textures or all of its DX11 features enabled. Basically nothing in this article will be relevant for the shipping version of BF3.

Not sure if you are accusing [H] of posting articles that are designed to be nothing more than trolling flamebait, but you've insinuated it without actually saying anything that will make you culpable.

Second part of your post is a good reminder to people who are getting excited, one way or the other, about card performance. Especially if they are going to get all knotted up between brands. This is a beta. They are collecting info about how the game runs. Changes are likely to be made before release. Then patches are likely to be made after release.

"stands up and cheers this post" :thumbsup:

Whats the excuse now why Hardocp's benchmarks once again are the total opposite of every other site out there. The old excuse was canned benchmarks vs real world gameplay.

Last I checked there is no canned bench for BF3.

I said it once and I'll say it again, Hardocp's biased, controversial, and more often then not, crap benchmarks are not worth reading, as is this thread.:thumbsdown:

Nothing wrong with [H]'s preview here. Read the article. It's all pretty basic on the procedures they use. If you have a legit complaint about them point it out. You're PO'd because your favorite brand didn't come out on top. Even though [H] states that there's a reason they did separate graphs for each card...

 

[RussianSensation]

The game is a beta - a DICE beta at that - it's not optimized, it's not even shipping with the highest quality textures or all of its DX11 features enabled. Basically nothing in this article will be relevant for the shipping version of BF3.

+100

 

[happy medium]

How can anyone expect consistency across different benchmarks from different sites.

There is one thing that is consistant on Hardocp's site, the fact that everytime there is a new game released or new card released, the AMD cards will somehow come out on top.
If there is a way they will find it, I garauntee it!

 

[skipsneeky2]

That Vr article is saying you need 1.3gb of vram for ultra at 1920x1080?


D: Dice/Ea needs to tweak their system or revise their recommended nvidia video card on the system requirements.

If what VR posted in that article is true,every nvidia before the gtx580 is almost useless for ultra unless you drop the settings or find a exotic 560ti or 2.5gb gtx570....

 

[Grooveriding]

I did my own looking into this and at 1080p ultra , I dont a see significant drop in my fps. Don't take my word for it

Considering the experience I had running a single overclocked 480 at 2560x1600 and then 1920x1200, I wouldn't want to see what it's like trying to play on ultra with a single 460.

I don't enjoy playing a slideshow.

http://vr-zone.com/articles/battlef...reenshots-and-hd-video-walkthrough/13625.html

'At 1080p Ultra settings (with everything graphics details set to maximum), we get an average of around 50fps for a single AMD 6970/GTX580, 80fps for a AMD 6990 (2 GPU crossfire) and 140fps for a AMD 6990 in Quadfire, which is actually quite similar to the one-year old Battlefield: Bad Company 2 on maximum settings. It also needed around 1.3GB of GPU framebuffer for the ultra preset textures. A side-by-side monitor test showed indistinguishable image quality from both AMD and Nvidia solutions (drivers at default settings), so we know that there is no quack3.exe foul play yet.'

Heck, I've seen posters on this forum saying they have to dial down settings on a single 580 @ 1080p. You need to get over whatever bothers you about BF3 being VRAM hungry.