Battlefield 3 Beta GPU & CPU Performance

[Guest]

Processors that only support 2 threads will take a hit as our Phenom II X2 560 delivered 42fps, almost 20% slower than a similarly clocked Phenom II X4 processor. However we were surprised by how well dual-cores performed when compared to other recently tested games.

This is a flawed conclusion since it looks as though the game is getting GPU limited at ~ 50 fps on your GTX 580 and the settings you used.

For a CPU benchmark you need to bench the game at a lower resolution and/or lower settings (medium settings) to make sure the benchmark is CPU limited rather than GPU limited. With the GPU limitation holding back frame rates on the quad cores it's impossible to tell how much faster the quad cores are over the dual cores CPU's. At lower graphic settings the Phenom II X2 may be much slower than 20% compared to the quad core Phenom II X4.

 

[Guest]

nice review but you should use a 1gb version of hd 5870 as most of us have 1gb version rather than a 2gb version.
and i am very sad that my hd 5870 is slapped by BF3 as it got only 27 fps on ultra at 1920x1200 i think turning aa off will increaee my rps atleast to 30+.
now i have some questions
1. in graphic settings menu which settings changes tesselation?
2. will my phenom 2 x6 2.7 ghz will perform as good as 3.3 ghz x6?
3. what is antialiasing post?

 

[Guest]

What setting was used on the ATI catalyst drivers? Were they set to HQ from the default Q?

 

[Guest]

WTF? Did Intel just sucked compered to AMD CPU, or what? I just made a big lol.

Intel i7 2600k = 300$
AMD Phenom II X6 1100T = 160$


How do you come to a tech site and make a comment like that? BTW, did you also notice the overclocked intel CPU's weren't much help? Any real techie could tell you the game relies more on GPU than CPU. Hence a single 6950 scoring 28fps on Ultra, but two 6970's get 98fps on HIGH. If a techspot writer or anyone else with a brain (besides myself), read your post and didn't *facepalm* I'd be real surprised... and disappointed.

 

[Guest]

Ohhh man, both AMD Phenom II X6 1100T and AMD phenom ii x4 defeat Intel i7 2600k
i wonder what the upcoming monster, amd eight cores bulldozer FX 8150p will do with this amazing game.



A whole 3fps more?! WWWINNING! Too bad they used a GTX 580 in that test, which costs more than your whole computer (Athlon II and 9600GT).

 

[DanUK]

I'm personally not too fussed about the current bugs in game as I think they will fix these (if you check out their twitter feed you can see they are personally replying to peoples tweets about bugs). Remember guys - this is a beta.

Also the CPU/GPU requirements seem reasonable enough for a game of this generation.

The thing I think that really bothers me is some of the functionality. The web client front end is just horrible, such a huge mistake imo. The kit customisation isn't user friendly and a step backwards from BC2... as is the way you can't choose what squad to be in (maybe this will change). Not being able to access the menu unless you're spawned is just crazzyyy stupid. I hope this is changed.. and the chat box too feels very primitive, i'm guessing this is just for beta only as it's so bad.

 

[en0nym0us]

RandyN said:
Why does Origin use ~150MB when doing nothing and having no in-game server is ridiculous.

That is because Origin is sending data reports of everything on and in your PC to EA.

 

[en0nym0us]

DICE is forcing you to use the new web-client in order to test the new functionality fully.

 

[Guest]

I just hope we'll see an improvement on frame rates in the final build. Fingers crossed

 

[Guest]

Fix C4 Glitch , Under the map Glitches, Weapon glitch

 

[Guest]

I'm sorry techspot, but you're losing readers with trash performance reviews like this.

I have a 2500k and a 6950 and run the BF3 beta with all settings maxxed including AA but with no motion blur @1080p and after running fraps benchmarks in half the games I play I have not yet seen a minimum fps below 30
My avg fps is more like 40 or 45 and the game is very very smooth at these settings

[H] agree with me on the 6950, so I just wanna say I ain't coming back here again

Oh yeah, and re: AMD vs Intel CPUs, who gives a **** if BF3 gives 4% better performance to AMD when every single last game else gives at least 10% to Intel

fanbois make me sick

 

[LNCPapa]

Was that the equivalent of a forum rage quit?

 

[1977TA]

Was that the equivalent of a forum rage quit?

I think so lol, I think TS shouldn't allow Guest (trolls) to post. But oh well...

 

[Steve]

LNCPapa said:
Was that the equivalent of a forum rage quit?

Haha I think it was

We always run the risk of losing readers when we post these performance articles, it’s just part of the game.

1977TA said:

Was that the equivalent of a forum rage quit?

I think so lol, I think TS shouldn't allow Guest (trolls) to post. But oh well...

If we did that I would lose a great source of comedy

 

[Guest]

This performance test is spot on for me. I have a GTX580 and at 1920X1200, the results are very similar.

 

[Guest]

i think ultra settings turns on or increase tessellation. as on high settings hd 5870 performs better than gtx 480, 560 ti and hd 6950 but when ultra settings are turned on hd 5870 falls behind all of them.

 

[Arris]

Guest said:
I'm sorry techspot, but you're losing readers with trash performance reviews like this.

I have a 2500k and a 6950 and run the BF3 beta with all settings maxxed including AA but with no motion blur @1080p and after running fraps benchmarks in half the games I play I have not yet seen a minimum fps below 30
My avg fps is more like 40 or 45 and the game is very very smooth at these settings

[H] agree with me on the 6950, so I just wanna say I ain't coming back here again

[H] also say

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.

In my own testing of 2 x 5850 I find that my results are just above that of a single 580 which is what I've seen in most other games, but I'm not going to stop reading any website just because their results are different from mines on a beta which has "inconsistent" performance making it "impractical to objectively compare performance between video cards".

Also you have to take into consideration that there is no set path benchmark with this beta. Outdoor areas are more hard hitting on performance than indoor on the Metro map. Maybe more of the test results taken by one site were from gameplay indoors compared to another. Until there is a set benchmark you are going to struggle to see consistency across results from different sites. These should really just be used as a rough guide to what we might expect from the full game.

 

[Steve]

I honestly did not find the game that inconsistent at all. Although I took the average of three runs (as we always do) each run was very similar in terms of performance. For testing we stuck to the outdoors as this is by far the most demanding aspect of the game and we tend to favor the most demanding scenes when putting these articles together. In any case I would say that guy was just trolling so mission accomplished.

 

[Arris]

I honestly did not find the game that inconsistent at all. Although I took the average of three runs (as we always do) each run was very similar in terms of performance. For testing we stuck to the outdoors as this is by far the most demanding aspect of the game and we tend to favor the most demanding scenes when putting these articles together. In any case I would say that guy was just trolling so mission accomplished.

That would explain the higher average reported by other benchmark reviews. I found that often in games a large amount of the play would occur in the underground higher FPS sections. HardOCP say that they played for 7 minutes, for me that would suggest that games would spend the majority of the time in the underground/indoor sections. Although I'm guessing that the guest probably just looked at the pretty bar graphs and didn't read the details.

 

[Guest]

The article says that there is little difference between a dual core and a similarly clocked quad, but that is untrue. I have a Phenom II x2 clocked at 4.1 Ghz and on OP Metro CPU usage is pretty much at 100%, while my GTX 580 is used at about 75-50% which a shows CPU bottleneck. Even so, the frame rates are very playable - low 30's at the worst, most of the time mid 40's.

On Caspian Border, however, my dual core is completely murdered. Even with the GTX 580 my frame rates drop into low 20's while the graphics card is utilised at about 50% !!! Do NOT try to run this game with anything less than a quad core! Phenom II x4 965 costs peanuts nowadays, and seems to be more than capable of running this game.

 

[Guest]

We are not playing the latest build. This beta is running one a build that is a few months old and is primarily for server load testing.

 

[Guest]

Yeah, there's no bench so it's inconsistent but [h] never see the 6950 drop below 30 fps in an 8 minute run which they claim to have iterated a couple times to confirm and I too never see the 6950 drop below 30 fps, and yeah, I've played the opening section of Metro quite a few times, so this ain't trolling, it's a total WTF ??? @ techspot, and between this and the Metro 2033 review which benched stuff about 50% higher fps than anyone else without explaining how or why I just lost my trust in you guys :/

sorry, but I ain't a troll, I'm a radge malaka who's bored of these balls

 

[red1776]

this and the Metro 2033 review which benched stuff about 50% higher fps than anyone else without explaining how or why I just lost my trust in you guys :/

Care to point out exactly where that was?

 

[Guest]

bc2 was a console port and bf3 is not.

its like the good old days folks when you just needed a normal cpu and a very good gpu.

 

[Guest]

you never needed the i5/i7 cpu's for the past 2-3 years, console ports were to blame.

my Q9650 at 4ghz is more than enough for BF3.

intel cought a break with i5/i7's.

good to see that BF3 is not a console port.