Battlefield 3
Battlefield 3 (DirectX 11)
Battlefield 3™ leaps ahead of its time with the power of Frostbite 2, DICE's new cutting-edge game engine. This state-of-the-art technology is the foundation on which Battlefield 3 is built, delivering enhanced visual quality, a grand sense of scale, massive destruction, dynamic audio and character animation utilizing ANT technology as seen in the latest EA SPORTS™ games.
Frostbite 2 now enables deferred shading, dynamic global illumination and new streaming architecture. Sounds like tech talk? Play the game and experience the difference!
Our Settings for Battlefield 3
Here is our testing run through the game, for your reference.
Initial results show us that the GTX 780 is performing very well, better than the GTX 680 and the HD 7970 and surprisingly close to the GTX Titan.
Frame times on the single GPU results are very consistent.
You can more easily see the average performance advantages of the GTX 780 over the GTX 680 and Radeon HD 7970 in this graph; the average frame rate is around 105 FPS on the GTX 780 and about 90 FPS on the HD 7970.
Frame time variance is low across the board – never breaching 1.5 ms.
As we increase the resolution the cards creep a bit closer and the GTX 780 sits more closely in the middle of the HD 7970 and the GTX Titan.
Frame times remain consistent with the single GPU results.
The average frame rate across the test run with the GTX 780 is around 65 FPS for BF3 at 2560×1600; the HD 7970 comes in behind it with a score of 57 FPS or so; the GTX Titan is around 75 FPS with a strong advantage.
There is hardly any frame time variance to be found here either.
As we creep into the world of triple display resolutions like 5760×1080, the separation between the cards is more important than before. Clearly the order goes Titan, GTX 780, HD 7970, GTX 680 though only the Titan is able maintain a frame rate over 40 FPS the majority of the test run.
Frame times do start to see a bit of fluctuation but still nothing dramatic with the exception of the spikes of the blue line representing the 2GB GTX 680.
All four cards are neatly spaced in terms of their 50th percentile frame rates – close the average results with the GTX 780 planted firmly between the GTX Titan and the Radeon HD 7970.
Frame time variance is still excpetionally low here at 5760×1080 with only the GTX 680 surpassing the 4 ms mark in our measurement.
Yay for Mini-Titan! looking
Yay for Mini-Titan! looking forward to live stream!
My golly AMD crossfire is
My golly AMD crossfire is still so screwed up AMD cheats so badly who can trust such a scamming con artist no wonder they need to bundle 3 or 4 games with their videocards to even sell a few of them.
Worse yet poor and blind AMD fanboys invade nearly every forum spewing an endless stream of lies and whining about nVidia prices when they can’t even buy a midrange AMD gpu, which explains why they go insane squealing about prices.
Get a freakin paper route or mow a few lawns AMD crybabies.
No AMD gpu’s are not good they are total crap compared to nVidia and nVidia’s massive software advantages and newly integrated game settings and upcoming streaming video to the handheld and PhysXand stable drivers and frame rate target and FXAA and far superior SLI and you name it.
When you still get the AMD corner mouse cursor bug and GSOD’s a unique AMD only sad crash I’ve had to put up with far too many times, WHY is the question.
I’ve had to waste about 20 days of my life helping idiots who bought AMD cards get the stupid things installed and running half crapped then they revert to turdville and the AMD fanboys squeals they didn’t do a thing to destroy stability.
OMG I hate them so much.
I should sue AMD for wasting human lives.
I think it’s odd you are
I think it’s odd you are complaining about ranting AMD fan boys but then you go off and become a ranting Nvidia fanboy.
I’m not a fanboy of either – i’m only loyal to the almighty Dollar (or dollar/performance ratio).
That being said, your argument about how bad AMD is seems like you did not read the article. The HD7970 Ghz edition is still the best bang for the buck for a single video card (no crossfire or SLI) and also includes 4 AAA title games. That’s worth a lot to most people that can’t afford $1000+ of video cards.
the 7970 doesn’t even run
the 7970 doesn’t even run half the games properly
How do you have the energy to
How do you have the energy to type out all that hate? I’ve been a loyal nvidia user for three cards, but couldn’t pass up a 7950 for $229 CAN. I was worried I’d regret it, but it’s posting some serious numbers. Overclocked it scored 3360 on the 3dmark11 and averaged 42fps on unengine Heaven. That betters the 670 and almost the 680. Considering those cards are going for $300+ I’d say I got a good deal.
You don’t have to pick one chip for life. You just have to try and find the best deal out there for what you’re willing to spend. I won this round with AMD, maybe next time I’ll go back to Nvidia, who knows?
You need to calm the fuck down and just buy what you want and stop shitting on whatever is competing with what you bought. Life is too short to be so angry.
If I just bought a Titan I’d
If I just bought a Titan I’d be pissed!
Honestly, if that was the
Honestly, if that was the case, then you would have bought the Titan for the wrong reasons.
Why the Titan is way more
Why the Titan is way more powerful, and dollar per dollar gets more bang for the buck!
I don’t like the price but it
I don’t like the price but it feels like they are forced @650 because the titan is 1000 and next years 880 will be more powerful then the titan, which will still cost $1000. I think the ball is in AMDs court right now and I think Nvidia won’t budge on price on anything unless AMD steps up big time.
You are exactly right. And
You are exactly right. And this is exactly what happened in the past with AMD’s video cards. There was a gap between the GTX 580 and GTX 680 when AMD had the HD6970 (or for those smart people, the unlocked HD6950) and for a while AMD’s prices were way too high because Nvidia had nothing to offer.
I was really ticked when the
I was really ticked when the Titan was released as I had just bought a factory OC GTX 680 4GB. I would have sunk that money into the Titan instead. But… it’s a good thing I didn’t sink my $ into a Titan. With the federal furlough going on until the end of the year, I’m having to do some belt tightening. Hopefully with the new fiscal year everything will be back to normal. I’ll probably wait until November and unload my GTX 680 for about 30%-40% off of what I paid for it and re-invest my money into a factory OC GTX 780. Thank you for the news release, I knew it was coming but its nice to see the performance numbers as well! Great job on the review!
Good review guys. Looking
Good review guys. Looking forward to your review video!
I’m confused about which
I’m confused about which drivers you were using for what. According to your test system page its 320.18, but you make continuous references that the data from the not-780 cards is being done using older drivers.
If that is the case you should specify which cards are using which drivers, for clarity if nothing else (were they just 320.14?).
Really, I just keep looking at the FC3 graph were the 780 is being shown outperforming the TITAN and going “wtf?”.
In FC3, that’s correct – and
In FC3, that's correct – and only because this latest beta driver we were given has special improvements for that game. We didn't have a chance to test Titan or 680 before publication.
Great review. Can we expect
Great review. Can we expect some Tri-SLI numbers from you guys?
Only about 3 hrs. until we se
Only about 3 hrs. until we se Ryan !
It`s getting to the point
It`s getting to the point that we may need the GPU`s to be in a separate box with a dedicated power supply/fans , eh ?
Nvidiaaaaaa.. What is with
Nvidiaaaaaa.. What is with these outrageous prices. Now when the Maxwell comes out i guess we should be expecting prices in 2K range.. Im losing my grip with Nvidia being a nvidia’s graphics card owner..Sigh AMD I wonder what will be on your side of things in the coming years.
So I’m looking at Skyrim sGPU
So I’m looking at Skyrim sGPU and on the frametime graph I’m seeing those nasty spikes on the geforces that look like going of the charts, and then I look at the percentile charts which don’t seem to reflect these spikes. I would think those are the sort of spikes that are noticable hitches during gameplay, no? Are the percentile charts averaging out the spikes or what? Am I missing something? wouldn’t be the first time lol
It would be nice if you’d add
It would be nice if you’d add a bar graph chart for ease of reading. I skipped to the end after the first benchmark just to see the conclusion. You need min fps at least in a chart (or min+avg), as I couldn’t care less about max which affects nothing for my game. It takes seconds to look at bar graphs, it takes minutes and is frustrating with a bunch of lines. I’m not saying remove those (maybe some like them), but you need a chart I can quickly see who had the best min/avg.
The current way you show the benchmarks is basically a big mess IMHO. I’m not talking about frametimes here, you clearly need lines for that to show. I’m talking the actual fps charts, those need to be done in bar graphs like most other sites do. They are very quick to read and get a quick picture of who is leading X game. You lose hits from people like me every review. Sure I can decipher all the lines, but I don’t have 10 minutes to do it for each chart. Instead I read the first page, the temps/noise page and the conclusion page. The charts are all just a PITA to me 🙁
Don’t get me wrong, I love the site, and when I have time I come back at some point (usually) to read more. It’s just easier to go elsewhere even if I don’t really want to.
I don’t think you understand
I don’t think you understand the graphics for FPS, which are pretty easy to see. It doesn’t show a number, but a chart of where the FPS was at each point in the 60 second benchmark. If you look at the lowest point in the graph, that is the minimum, if you look at the highest point, that is the maximum. This method gives you a clearer picture of what actually is taking place. Especially with minimums. In the bar graph charts, you don’t know if the minimum is a small drop at the load of a game, or if the game gets near that point often. This method makes it clear.
Can you please tell me the
Can you please tell me the exact settings used in your heaven benchmark.
The extreme does not match up with results I’ve seen if thats 1920×1080 w/ 8xAA
I just ordered the 780 and im
I just ordered the 780 and im kind of geeking out waiting for it. Lol just kidding my question is everyone says this card is not good for a single 1920×1080 monitor but what about people like me who have the asus 144hz monitor and want to play games at 120fps or play games in 3d? Wouldnt this card be perfect for that?
Yeah I am with you, currently
Yeah I am with you, currently ordering a 780 but debating whether to get a high res monitor with 60fps or a 1080p with 120, its the last thing I need to complete my build.
Is it faster than all but
Is it faster than all but Titan? Yes. Does it overclock well? Yes, easily going beyond Titan in many games, according to Hardocp. Does it use less power than 7970GE? Is it quieter? Yes. Is it 35% less than Titan? Yes.
Should it be 60 dollars less? Yes, and it will be soon. DO most high end buyers give a %@$#$% about 60 bucks? I doubt it.
I don’t think you use a
I don’t think you use a particularly good benchmark run for Skyrim. It takes place primarily in Whiterun, which is more CPU bound in my experience. A better choice for GPU benchmarking would be the forest areas either around Riften or, preferably, Falkreath. These are the most GPU bound areas of the game.