[H] Battlefield 4 Video Card Performance and IQ Review

[wand3r3r]

HardOCP has a Battlefield 4 GPU / IQ Review up.

It clearly looks to favor AMD cards. That's not very surprising considering AMDs involvement.

From a gameplay performance perspective, the AMD Radeon HD 290X came out on the top of the heap, finding itself playable at 2560x1600 with all in-game settings set to maximum values and 2X MSAA enabled. The GeForce GTX TITAN followed by being playable at the same settings, however, it produced an average frame rate about 10FPS below that of the R9 290X.

Taking a step down from the top two performers, the AMD Radeon R9 290, AMD Radeon R9 280X and NVIDIA GeForce GTX 780 were all playable at 2560x1600 with all in-game settings set to maximum values and low FXAA enabled. The GeForce GTX 770 was able to play at 1920x1080 with all in-game settings set to maximum values and 2X MSAA enabled.

Rounding out the lineup, both the NVIDIA GeForce GTX 760 and the AMD Radeon R9 270X was capable of playing at 1920x1080 with all in-game settings set to maximum values and low FXAA enabled.

From an overall gameplay perspective, it seems that the Frostbite 3 engine has gone a long way to prevent significant sudden frame rate dips from impacting the gaming experience. Playing Battlefield 3, these frame rate dips (for example, when taking out a tank) could easily render your ability to escape from the situation futile. With Battlefield 4, these dips do not feel like these bring action to a halt.

What was more disconcerting was the inconsistent performance between AMD and NVIDIA cards. Overall, AMD cards tended to perform better than the NVIDIA counterpart at the particular price point, with the exception of the GeForce GTX 760 vs. the Radeon R9 270X.

With each NVIDIA card, we observed frame rates to be far more varied than AMD based cards over the course of playing the game with them. It almost seems that the performance concerns that we had with NVIDIA cards during our Beta evaluation have only been partially fixed at this point.

http://www.hardocp.com/article/2013/11/17/battlefield_4_video_card_performance_iq_review

 

[Grooveriding]

Interesting that they didn't include the 780ti, although the 290X is 20% faster than a Titan in Battlefield 4 multiplayer there, so the 780ti would not of equalized things for nvidia with AMD having that much of a lead. Not to mention it costs $200 more...

I hope they revisit this next month when Mantle comes out in Battlefield 4. I want to see DX11 AMD vs DX11 Nvidia vs Mantle AMD.

 

[ozzy702]

I'm also interested in how game and driver improvements change the "battlefield". It appears that AMD has a leg up on NVIDIA with BF4 and with mantle that lead should increase. I wonder how much NVIDIA will be able to keep up, especially since I'm rocking sli 780's and BF4 is one of the only games I have enough free time to play on a regular basis.

 

[Attic]

Interesting that they didn't include the 780ti, although the 290X is 20% faster than a Titan in Battlefield 4 multiplayer there, so the 780ti would not of equalized things for nvidia with AMD having that much of a lead. Not to mention it costs $200 more...

I hope they revisit this next month when Mantle comes out in Battlefield 4. I want to see DX11 AMD vs DX11 Nvidia vs Mantle AMD.

Yea, agree. Unfortunate that we'll have to wait for the Mantle effect since it is a driver for considerations between AMD vs nVidia and now is a big buying season.

I played BF4 on a friends 290x at 1920x1080x3 Eyefinity. Was not impressed with the graphics TBH. It looked great, but didn't wow me at all. I'm more of an Alice Madness returns type or Steam Punk graphics whore who would like additional shadow or texture detail there. That being said BF4 did have stunning draw distance and detail and in eyefinity res there was a definitely "holy cow" experience above what a I think a single display would have delivered. The 290x paired with a i950 processor didn't maintain 60fps.

 

[imaheadcase]

/sees Hardocp link. Closed tab.

 

[GaiaHunter]

Interesting that they didn't include the 780ti, although the 290X is 20% faster than a Titan in Battlefield 4 multiplayer there, so the 780ti would not of equalized things for nvidia with AMD having that much of a lead. Not to mention it costs $200 more...

I hope they revisit this next month when Mantle comes out in Battlefield 4. I want to see DX11 AMD vs DX11 Nvidia vs Mantle AMD.

It seems it wasn't by choice.

(Kyle’s Note: Let’s address the elephant in the room; the lack of the GeForce GTX 780 Ti which we fully reviewed here and here over the last ten days. Yes, we would have very much liked to have included the GTX 780 TI in this article but resources and logistics kept us from that. That said, given the results we saw at 2560x1600 among the high-end card set, we highly doubt that the conclusions we reached would have changed at all considering the much less expensive Radeon R8 290 is our suggested card for BF4 at 2560x1600 resolution.


We have spent a lot of time with BF4 simply learning how to properly test for our readers. We will continue to test BF4 in a multi-player only environment because we understand that is how the huge majority of our readers will be using it. BF4 is being included in our regular testing suite so you will see a lot of coverage going forward. Surely you will see the GTX 780 Ti and BF4 testing here on [H]ard|OCP soon, both at multi-display resolutions and in multi-GPU configurations. We also have some specialty articles focused around BF4 that we will be publishing as well. So stay tuned on that front.)

 

[chimaxi83]

/sees Hardocp link. Closed tab.

Is that because Nvidia is in second place in the review? :awe:

 

[96Firebird]

The fact that the 760 beats the 770 in the apples-to-apples leads me to believe there is a LOT of variation with regards to the "toughness" of the benchmarks for each of the cards. I see they touch on it there, but that speaks a lot to me...

 

[VtPC83]

Is it just me or does HardOCP tend to favor AMD cards in their benchmarks, no matter what other sites have published? I always tend to find HardOCP stating AMD cards are better the majority of the time...

 

[Grooveriding]

Yea, agree. Unfortunate that we'll have to wait for the Mantle effect since it is a driver for considerations between AMD vs nVidia and now is a big buying season.

I played BF4 on a friends 290x at 1920x1080x3 Eyefinity. Was not impressed with the graphics TBH. It looked great, but didn't wow me at all. I'm more of an Alice Madness returns type or Steam Punk graphics whore who would like additional shadow or texture detail there. That being said BF4 did have stunning draw distance and detail and in eyefinity res there was a definitely "holy cow" experience above what a I think a single display would have delivered. The 290x paired with a i950 processor didn't maintain 60fps.

Surely. That sounds more like an artistic preference though. Tech wise Battlefield 4 is incredibly impressive. BF4 and Crysis 3 are the pinnacle of what is available right now on the PC. Multiplayer game wise nothing is like BF4 as well.

DICE managed to make some good improvements to the circumstances that would cause slowdowns in BF3 with BF4. As a whole BF4 is a noticeably more demanding game than BF3 was; across the board my performance is about 25% lower than it is in BF3. But things that caused large and noticeable frame drops in BF3 like large-scale destruction or explosions close to where you are in a multiplayer game don't give the big frame drops they would in BF3.

I'm not surprised with the lead AMD has here. I've seen it in every other multiplayer review done except a fringe one from some random no-name site I saw posted on [h] forum which appeared completely broken.

 

[BallaTheFeared]

The fact that the 760 beats the 770 in the apples-to-apples leads me to believe there is a LOT of variation with regards to the "toughness" of the benchmarks for each cars. I see they touch on it there, but that speaks a lot to me...

I'd argue that MP should be reserved for CPU testing and only over very long sampling periods.

SP is the right way to go for GPU testing.

 

[ozzy702]

Is it just me or does HardOCP tend to favor AMD cards in their benchmarks, no matter what other sites have published? I always tend to find HardOCP stating AMD cards are better the majority of the time...

It seems like they usually say that AMD cards provide a better price/performance ratio. I don't feel that they lean AMD in regards to performance at least not in the same way that many other sites lean to the green team.

I bounce back and forth between NVIDIA and AMD and don't really care as long as I have a good end user experience. My single 7970 was a great setup but the crossfire 7970 even after the frame pacing drivers sucked so I jumped ship to SLI 780s and have been pleased with the overall experience. I may jump back to AMD next time around since it appears that AMD has really been kicking ass with the 290/290X but I won't jump until 2015 when 20nm is fairly mature.

 

[thilanliyan]

Is it just me or does HardOCP tend to favor AMD cards in their benchmarks, no matter what other sites have published? I always tend to find HardOCP stating AMD cards are better the majority of the time...

Really? I always figure [H] to be nV biased, but they don't LIE, so if an AMD card is soundly beating the equivalent nV card, they're going to say as much. But if the cards were equal, they would favour the nV card, at least IMO.

Personally, I wasn't a fan of their testing methods before but with all the "cheating" happening in canned benchmarks, I think their play through style is more indicative of what the average user will experience, and so have grown to like their methods more.

 

[Kenmitch]

Is it just me or does HardOCP tend to favor AMD cards in their benchmarks, no matter what other sites have published? I always tend to find HardOCP stating AMD cards are better the majority of the time...

I'm not sure but I think they go back and forth all the time.

They don't go the canned benchmark way on their reviews.

Evaluation Method



We evaluate what each video card configuration can supply us in terms of a playable gaming experience while supplying the best culmination of resolution and "eye candy" graphical settings. We focus on quality and immersion of the gameplay experience rather than how many frames per second the card can get in a canned benchmark or prerecorded timedemo situation that often do not represent real gameplay like you would experience at home.

I think it's the biggest reason why their results don't line up with other reviews. It does make some sense why they benchmark this way as a person will most likely play the game and not the included canned benchmark.

 

[Grooveriding]

I've seen people call [H] nvidia biased and then AMD biased, same way I've seen people call Anand AMD biased and then nvidia biased. Just look at that trainwreck about the the AT 290 review, flip-flopping galore from many sources of these bias claims... :sneaky:

Apart from a few very obvious whacky fringe 'review' (shill) sites, these guys give their own subjective views in tandem with solid benchmark results. If they were displaying obvious bias consistently either way and giving unfair results, they'd likely no longer be receiving hardware from the vendor they were sandbagging against...

Most claims of bias just = 'I don't like the results cause my favourite vendor is losing benchmarks'

We all knew going in (I hope) AMD would take BF4.

 

[krumme]

I'd argue that MP should be reserved for CPU testing and only over very long sampling periods.

SP is the right way to go for GPU testing.

Disagree. People play multiplayer. Benchmarking should reflect user behavior. Sp gpu benchmarks is more correct but ultimately useless unless it reflects mp bm.

 

[blastingcap]

Is it just me or does HardOCP tend to favor AMD cards in their benchmarks, no matter what other sites have published? I always tend to find HardOCP stating AMD cards are better the majority of the time...

This is only a problem if they are incorrect. What if AMD *are* better the majority of the time (and by that I think you mean that they say that AMD is better in price/perf in the way that they measure it... obviously TITAN is better than anything if you need DP compute for instance)?

 

[Teizo]

Why are they benchmark reviewing a game still in Beta? Wait until all the patches are out and the game is properly optimized.

 

[taq8ojh]

I'd love to read the article, because I am just about to hit the upgrade button, but omg, whoever thought white text one black background was good idea should be kicked in the balls. I am dizzy after five seconds of trying to focus

 

[Meekers]

The fact that the 760 beats the 770 in the apples-to-apples leads me to believe there is a LOT of variation with regards to the "toughness" of the benchmarks for each of the cards. I see they touch on it there, but that speaks a lot to me...

They test each game at maximum playable settings, which are lower for the 760 than the 770.

Edit: Whoops, just saw the apples to apples and the 760 did get slightly higher. Odd.

 

[BallaTheFeared]

Disagree. People play multiplayer. Benchmarking should reflect user behavior. Sp gpu benchmarks is more correct but ultimately useless unless it reflects mp bm.

MP is too inconsistent, and the engine is exactly the same as SP.

Not sure why you'd want them to test graphics performance in such a variable situation of cpu necking and randomness.

Not that I think it will matter much, but evidently the 760 is faster than the 770 according to H.

 

[chimaxi83]

MP is too inconsistent, and the engine is exactly the same as SP.

Not sure why you'd want them to test graphics performance in such a variable situation of cpu necking and randomness.

Not that I think it will matter much, but evidently the 760 is faster than the 770 according to H.

You seriously don't see the merits of MP testing in Battlefield?

 

[BallaTheFeared]

Not for graphics, for CPU, yes.. Very much yes.

 

[BrentJ]

Interesting that they didn't include the 780ti, although the 290X is 20% faster than a Titan in Battlefield 4 multiplayer there, so the 780ti would not of equalized things for nvidia with AMD having that much of a lead. Not to mention it costs $200 more...

I hope they revisit this next month when Mantle comes out in Battlefield 4. I want to see DX11 AMD vs DX11 Nvidia vs Mantle AMD.

Testing was completed when the 780 Ti came out, it didn't make it in the review because it was released too far into the review schedule.

In future reviews, and our future looks at BF4, including SLI and multi-display, 780 Ti will be included.

 

[BrentJ]

Is it just me or does HardOCP tend to favor AMD cards in their benchmarks, no matter what other sites have published? I always tend to find HardOCP stating AMD cards are better the majority of the time...

Before the 290X launch we were called NVIDIA Biased, because the real-world results favored that at the time.

It goes both ways, and when we are called both NVIDIA and AMD biased, we know we are doing things right.