BattleField 4 Performance AMD vs Nvidia

Performance Testing

BattleField 4 Performance Comparison

Introduction

BattleField 4 is the poster child for the next generation of consoles. With a relative dearth of AAA titles it has fallen to DICE to wow us with the latest iteration of their first person squad shooter.

As much as we might champion the cause of the PC Gaming Master Race™ we're very aware that in these financially difficult times ever more companies are developing primarily for the console market, and then porting to the PC. Sadly the graphical capabilities of consoles have long been far behind that of the PC and so we've been stuck with games which didn't take full advantage of the tremendous hardware available to us.

With both the Playstation 4 and the Xbox One sporting relatively powerful GPUs, games are finally looking as good as we'd hoped and this could only be good for PC gamers. Thankfully BattleField 4, whilst a bit bug-ridden to play, looks the business and has been released on all major formats, so it's the perfect opportunity for us to bring together the top three graphics cards from both AMD and nVidia and see which ones give us the best performance.

For our testing we'll be running on our LGA2011 rig that we usually use for graphics testing, which consists of the following :

Intel Core i7-3960X @ 4.6GHz
ASUS Rampage IV Extreme
Corsair Dominator Platinum
Corsair AX1200i
Corsair Neutron GTX
Corsair H100i
Windows 7 x64

To this we'll be running the reference versions at reference clocks of the AMD R9 270X, 280X, 290 and 290X, and the nVidia GTX770, GTX780 and GTX780Ti . There is no doubt that the Frostbite 3 engine looks the business, with good texture detail and some excellent anti-aliasing and lighting effects. DUring all tests the game was sex to maximum settings at 2560x1440 resolution.

R9 270X                                                                        R9 290

BattleField 4 Performance Comparison R9 270X     BattleField 4 Performance Comparison R9 290  

R9 290X                                                                      GTX770

BattleField 4 Performance Comparison R9 290X      BattleField 4 Performance Comparison GTX770  

GTX780                                                                  GTX780Ti

BattleField 4 Performance Comparison GTX780     BattleField 4 Performance Comparison GTX780Ti  

Performance

We were expecting the AMD cards to perform better, especially as so much focus and optimisation in development will have gone towards that particular architecture. To some degree this is true with both the 290 and 290X out-performing the GTX780 and the 280X bettering the GTX770 by 5 FPS average.

There isn't much surprise at the top and bottom of the chart though, with the GTX780Ti maintaining its role as the fastest card around but the most surprising result to us was the 270X with its relatively small price tag putting out those kind of FPS numbers with the game both maxed AND in 2560x1440.

BattleField 4 Performance Comparison  

So there you go. If you wondered how BattleField 4 would perform on your system, or which graphics card you should upgrade to, hopefully this will help. Everything but the 270 handles it smoothly at this ridiculously high resolution and level of detail. A reduction down to 1080P will enable even the 270 to perform well, which shows that as good as the very latest consoles are, PCs will always have plenty of power to spare.

 The performance award has to go to the graph leader here the GTX780 Ti

 Gold award deserves to go to the R9 290 the performance for the price puts the 290X to shame.

 The VFM award has to go to the 270X, even with the high res and max details it still put out some awesome numbers, a simple resolution drop to 1920x1080 would make a massive difference and we all know if you have a £140 GPU you wont be running a high resolution monitor like this.

Are you enjoying playing BF4? Did these results surprise you? Let us know in the OC3D Forums.

«Prev 1 Next»

Most Recent Comments

02-12-2013, 03:54:44

tinytomlogan
http://www.overclock3d.net/gfx/artic...161939365l.jpg

We put the R9 270X 280X 290 290X and the GTX770 GTX780 & GTX780 Ti head to head in BF4 to see what all the fuss has been about.


Continue Reading

02-12-2013, 04:05:46

ShaunB-91
For how good the game looks it's surprising how well GPU's can handle it. I was biting my nails and clenching my arse when I first got the beta hoping that my 7970 could run it and was shocked/eroused/chuffed when it was getting a connstant 60fps, even jumping to the 90's on occasions, and it seems to run even better on the full release of the game.

The only issue now is the bugs!

02-12-2013, 06:02:12

Pendragon
I'm running a pair of 7970's, BF3 performance is flawless. at times the game hits the internal game cap of 200fps. I was a little surprised by the drop in performance playing BF4 MP as it visually looks much the same as BF3. Granted the BF4 single player is the most amazing looking game I have seen to date but the MP game not so much. I would guess that I have lost around 50fps moving from BF3 to BF4 with graphics that don't look all that different.

The elephant in the room is Mantle, will it be the most amazing thing ever? A real leap in gaming performance? Time will tell but if you are building a gaming rig for Battlefield fitting a nVidia right now would be a tough decision...

02-12-2013, 06:06:09

ShaunB-91
I'd personally just get a 780/Ti. I was debating getting a second 7970 at the weekend but I don't really need it to be honest. I suppose if I played a lot of games it might be worth it, Far Cry 3 seems to be stuck at the 30fps mark and Arkham Origins at 21. I thought about a 780Ti but it's too much for me to spend really for the amount I'd use it.

02-12-2013, 06:13:53

Pendragon
If your PC can take another card, it would be a very cost effective upgrade. Modern cards scale very well, a pair of 7970's is still faster than a single 780Ti.

A 280X costs what? 240 quid, buy a 780Ti get poorer performance for an extra cost. Not knocking the 780Ti it's a great card but if you already have a 7970 why not fit a second one???

02-12-2013, 06:19:06

ShaunB-91
Main reason I decided not to go for a 780/Ti is because then I'd have a GPU I'd need to sell which I've owned for about three months. I may consider getting a second 7970, maybe after Christmas. My only worry is there aren't as many around now and I need to make sure I get the exact same card. Overclockers don't have it were I bought it from now and Scan say sold out. :|

02-12-2013, 06:27:38

SieB
I'd just wait for custom coolers on the 290.

A 290 is 330 and a 780Ti is 550.
The thing it comes down to for me is if the ~10fps extra from a 780Ti is worth an extra 220? Personally, I don't think so. Custom cooled 290s are going to be more expensive than reference ones, but not that much and they are still going to be massively cheaper than a 780Ti.

It will be a toss up between the 780 and 290 for me but with Mantle coming it will most likely be the 290 with a DCU II for me.

02-12-2013, 06:41:39

Excalabur50
If this useless game would stop crashing every five minutes we might get to enjoy the stunning visual maps and awesome sound that is BF4 but until then it's just frustration and total rage I'm just so glad I wasn't stupid enough to buy premium.

02-12-2013, 07:04:07

Xrqute
Quote:
Originally Posted by ShaunB-91 View Post
Main reason I decided not to go for a 780/Ti is because then I'd have a GPU I'd need to sell which I've owned for about three months. I may consider getting a second 7970, maybe after Christmas. My only worry is there aren't as many around now and I need to make sure I get the exact same card. Overclockers don't have it were I bought it from now and Scan say sold out. :|
To my knowledge bud a 7970 with Xfire with a 280x no issue.

02-12-2013, 07:14:23

TommyGunn123
There's the 270 results but no 270x results, jussayin'

02-12-2013, 07:14:52

doombadger
Quote:
To my knowledge bud a 7970 with Xfire with a 280x no issue.
Yes this is my understanding as well. Also, I believe Crossfire works between vendors too so a Sapphire will work with an Asus, EVGA or any other just fine. If I'm wrong I'm sure someone will correct me.

02-12-2013, 07:42:51

ShaunB-91
But if I get the exact same cards then they'll match and look much better.

02-12-2013, 08:42:08

Tripp
one thing thats going to change the scores is mantle, thats one reason am surprised this is being done now an not in a fortnight

02-12-2013, 08:49:38

SuB
I assume these numbers come from the single player campaign?

When mantle comes in, I think it might be worth ditching the single player and going for multiplayer figures, that's what really matters and it *does* run differently, (oftentimes slightly better) than single player.

02-12-2013, 08:58:02

Watsyerproblem
Quote:
Originally Posted by Tripp View Post
one thing thats going to change the scores is mantle, thats one reason am surprised this is being done now an not in a fortnight
I imagine tom is doing this to get data before and mantle and compare it to upcoming mantle results when it is patched in.

he also gets 2 articles out of it, both of which will be relevant for pc gamers/enthusiasts in different ways.

02-12-2013, 09:56:37

ermacos
My card is ASUS direct cu2 gtx780, that means its the same as 290?

I am going to have it tomorrow (90%) in 2 days (10%), depends the currier.

I will sli it of course, 2 of them, but not now.


ps: I hope I did a nice choice! I wanted a low watt, high performance and cool card. I think that ASUS direct cu 2 OC 780, consumes a bit less power than 280x?

02-12-2013, 10:06:46

Tripp
Quote:
Originally Posted by Watsyerproblem View Post
I imagine tom is doing this to get data before and mantle and compare it to upcoming mantle results when it is patched in.

he also gets 2 articles out of it, both of which will be relevant for pc gamers/enthusiasts in different ways.
good point mate

02-12-2013, 14:47:51

notcool
But I thought you needed 10GB of vram for this, oh wait... hopefully this review will stop alot of the vram scaremongering that's going on. System cruncher you say hah, good one.

02-12-2013, 15:11:39

SeraNoxa
You should do the same on windows 8.1 just to show what extra benefit the cards will gain from that.

02-12-2013, 20:21:12

qhfreddy
Nice results from both sides...

Also good to see that DICE got the graphics optimizations done pretty solidly on release.

03-12-2013, 04:31:22

Xrqute
Quote:
Originally Posted by ShaunB-91 View Post
But if I get the exact same cards then they'll match and look much better.
Well if you can't get the same. Take off the coolers, put them on water and you'll never know the difference they'll just look like two Wc'd cards.

03-12-2013, 04:50:40

SuB
Quote:
Originally Posted by notcool View Post
But I thought you needed 10GB of vram for this, oh wait... hopefully this review will stop alot of the vram scaremongering that's going on. System cruncher you say hah, good one.
It's funny, because the requirement is 3GB, and I will 'scare' you by saying I see the game use 2.5GB on a nightly basis Good one eh?


Quote:
Originally Posted by qhfreddy View Post
Also good to see that DICE got the graphics optimizations done pretty solidly on release.
This is quite funny, they are STILL plagued with issues, crashes and a bunch of other stuff. Systems that didn't work, were stable and then now crash in almost every game... EA screwed this release up. You'll see from my previous posts I've been willing to give them the benefit of the doubt as my experience was largely un-touched by the issues. But they are starting to rear their ugly head with fellow players now.

As for optimisation, that's a complete joke. There isn't THAT much visual improvement over BF3 to be fair and it's no where near as optimised as it could be.

While I'm happy overall with this game, there is still a lot left for DICE to do, while EA push push push for closer release dates etc....

Still think we need Multiplayer benches, the single player isn't really what I'd consider 'representative' of the final product.

03-12-2013, 05:01:31

ShaunB-91
Some issues should be patched up today, or is it tomorrow? O.o

http://www.gamespot.com/articles/bat.../1100-6416509/

03-12-2013, 05:06:59

barnsley
If you get a 270x/270 you don't need to get BF4 as the asus ones (not sure about the others) come with it.

04-12-2013, 15:38:51

Wolvie87
And here's me playing on a 6970

09-12-2013, 07:52:46

ShaunB-91
Quote:
Originally Posted by Wolvie87 View Post
And here's me playing on a 6970
What's it like?

09-12-2013, 10:05:49

Wolvie87
Plays in Custom> High with a few settings on ultra, smooth as a baby's bum

09-12-2013, 10:38:57

xRavaged
Quote:
Originally Posted by Wolvie87 View Post
And here's me playing on a 6970
Ha, and I'm here stuck with my 6950...
Weird how to can run fairly smooth high in multiplayer, only issue is when zooming in on recon, can definitely tell the card is struggling, poor frame rate....
Haven't testing in campaign properly tho, so probably only runs on medium ?

I need of an upgrade!

09-12-2013, 10:48:23

SieB
Quote:
Originally Posted by xRavaged View Post
Ha, and I'm here stuck with my 6950...
Weird how to can run fairly smooth high in multiplayer, only issue is when zooming in on recon, can definitely tell the card is struggling, poor frame rate....
Haven't testing in campaign properly tho, so probably only runs on medium ?

I need of an upgrade!
Even my 7950 gets frame dips when zooming in with sniper rifles scopes

I read somewhere that it's because even though you are looking down the scope that it still renders the map as if you are unscoped, plus it has to render what you are looking at down your scope. So basically your GPU is rendering more when scoped than when unscoped.

09-12-2013, 11:02:58

xRavaged
Arhhh so the GPU is working twice as hard rendering :L
Either way, I hate my XFX card, cant really overclock it to unlock shaders or anything... Don't think I could even crossfire it... No bridge connection?

12-12-2013, 20:45:18

SparkleDJackson
What the flip tom. You didnt sign out with the invisble pen click. I rewatched the ending to make sure I wasnt not seeing things

12-12-2013, 20:55:03

cooperman
Well this pretty much makes the R9 290X a GTX titan ---> a over overpriced pointless (to gamers) card that makes a GTX780/R9 290 look cheap....

12-12-2013, 22:11:06

Excalabur50
Th one thing I don't understand is why do all reviews focus on single player mode when the real test of performance is multiplayer it's like driving your car down the road where the speed limit is posted at 30mph or 60kph versus driving it on a race track where the speed is unlimited, total chalk and cheese scenario

13-12-2013, 03:33:58

Warchild
Because MP is too random.

You cannot compare gfx cards or benchmark MP when even using the same gpu, the FPS will vary from map to map, or even the same map.

13-12-2013, 04:08:29

Chopper3
I'd have liked to have seen the results when played at 1920 - I went 120hz rather than 2560 so care more about that aspect - but I do appreciate TTL only has so many hours in the day to do this sort of stuff.

13-12-2013, 04:08:45

SuB
Quote:
Originally Posted by Warchild View Post
Because MP is too random.

You cannot compare gfx cards or benchmark MP when even using the same gpu, the FPS will vary from map to map, or even the same map.
while this is true, battlefield is designed for multiplayer, and as such, it's pointless seeing singleplayer figures because that's not the usual usage scenario of the game...

It's a good way to get solid/repeatable results, granted but.. in my opinion it's still not the 100% correct thing to show. When the shit hits the fan in multiplayer you need to know your titan is gonna cut the mustard...

19-12-2013, 05:18:55

Preed
I'm sticking to my ideals and not buying anything from Nvidia since I don't like their way of doing business a'la Apple. So even if the 780Ti was half the price of a 290X I would still not buy one..

19-12-2013, 05:30:00

ShaunB-91
Can Tom even survive long enough in BF4 multiplayer to get some scores, I mean...the respawn point is pretty empty and quite so he'll be on some good high fps there.

The multiplayer is the proper task but it's way to random, you'd have to test for hours, one map (from my experience) could throw a good 80-90fps at you and your there with tight jeans, next round/map your at 40.

19-12-2013, 09:44:51

Chopper3
Public multiplayer runs are inherently a bad benchmark - if games developers wanted they could setup benchmark-specific servers with scripted and predictable 'other players' that reviewers could test against but I'm not aware of them doing this and in lieu of these I'd rather not see wildly-varying benchmarks.
Reply
x

Register for the OC3D Newsletter

Subscribing to the OC3D newsletter will keep you up-to-date on the latest technology reviews, competitions and goings-on at Overclock3D. We won't share your email address with ANYONE, and we will only email you with updates on site news, reviews, and competitions and you can unsubscribe easily at any time.

Simply enter your name and email address into the box below and be sure to click on the links in the confirmation emails that will arrive in your e-mail shortly after to complete the registration.

If you run into any problems, just drop us a message on the forums.