Battlefield 4 Beta Performance: 16 Graphics Cards, Benchmarked

[_BigHead_]

On page 8 of this article, the 3rd graph is labeled incorrectly. Great article, i was pleased and quite surprised when i saw it!

 

[cyan1d3]

It would be interesting to see some of the AMD 5xxx and 6xxx cards (5870 and 6970 maybe) and some of the Nvidia 5xx cards as well. Not everyone is rocking a 7xxx AMD or 7xx Nvidia

 

[goleksy]

Looks like I should be ok, at least for now, with my 3770k @ 5GHz and 670 4GB SLI at 1150/6400. I'll be driving a 1600p display though, I'm sure I'll be ok.

 

[spookyman]

I wonder if my 4960X with 2 Titan's in SLI will have a problem running this game?

 

[Neiromaru]

When I read the title I thought they were somehow testing it on a single system with 16 graphics cards running in some sort of hyper-SLI...

 

[Steveymoo]

So, I will need to upgrade from GTX 460 SLI huh? Or will iiii?

 

[vibol03]

These numbers will improve greatly when the release come out!

 

[n3cw4rr10r]

Nice to know my 7970 is good for another couple years ateast.

 

[horaciopz]

I'm looking at this article and love what i see. I wont need a 500$ ++ GPU to play smoth like Crysis 3. I have a problem, my HD 6950 2GB struggles to maintain 30fps (dips below 20 sometimes) and barely can go over 35fps. Shoulndt my GPU be at performance level of HD7850 2gb? Shoudlnt I be able to play on average 40fps on High preset @ 1080p? I mean, do i really need to upgrade? :/

 

[06yfz450ridr]

im getting 60-65 on my 7970 with 4xmsaa with wierd occasional drops @1200/6260

if i run 2xmsaa on ultra every thing maxed 80+ fps and doesnt dip as bad

Just sucks playing that crappy domination map where evryone camps the roof every game

 

[caamsa]

Any plans to do some crossfire testing with two 7770's

 

[cleeve]

I'm looking at this article and love what i see. I wont need a 500$ ++ GPU to play smoth like Crysis 3. I have a problem, my HD 6950 2GB struggles to maintain 30fps (dips below 20 sometimes) and barely can go over 35fps. Shoulndt my GPU be at performance level of HD7850 2gb? Shoudlnt I be able to play on average 40fps on High preset @ 1080p? I mean, do i really need to upgrade? :/

It's always hard to diagnose a remote system - no Idea what your CPU is - but download the patches as they come and lower detail as necessary.

 

[InTheCity]

"However you felt about Battlefield 3, you'll probably feel the same way about Battlefield 4". In that case, I'll avoid it. I played all Battlefield games prior to Version 3 and actually felt if I tried hard enough I could come in the top 3 at least 80% of the time. No such luck with BF3, I never came out on top in one and one gun battles, felt cheated out of headshots and just could never understand how I was getting my ass handed to me every time. That game was inherently flawed and my connection here is 100mbps, with almost no server lag. If BF4 is more of the same, then BF4 is not the game for me.

 

[cleeve]

Page 8, @frame time variance table, title is "Metro: Last Light Frame Time Variance".

Doh!
Thx, fixed!

 

[ryder553]

Will the R9 280X be better than a 7990?

 

[ifreestylin]

Would have loved to see the FX 6300. Everyone I know including me that owns an FX 6300 is getting bad performance.
My cousin who has the i5 3570k at stock stats between 45-60fps on ultra with AA off. In the GPU end, we all have either 7850/7870 also, my fx 6300 is Over-clocked to 4.6ghz and no matter what settings or resolution I use, my fps doesn't changed at all! (19-40) if it's a good day.

P.S Windows 8

 

[horaciopz]

Will the R9 280X be better than a 7990?

Actually the R9 280x is an HD7970 ghz edition, you should spect to be at the same range, so 2 R9 280x = HD 7990

It's always hard to diagnose a remote system - no Idea what your CPU is - but download the patches as they come and lower detail as necessary.

I know, my processor is an I5 3570k overclocked (i have to update my sing). It runs at 90% while im gaming Bf4 and the GPU is just working fine.

 

[mauller07]

if these benchmarks were taken before todays cpu and directx performance patch then they are irrelevant as the patch has improved performance dramatically across the board.

 

[whimseh]

I guess they forgot about GTX 680's

 

[ojas]

I guess they forgot about GTX 680's

Maybe because the 770 is a re-branded, slightly overclocked 680?

 

[cleeve]

I guess they forgot about GTX 680's

Maybe because the 770 is a re-branded, slightly overclocked 680?

Bingo.

 

[fulle]

I would have liked to see a FX-6XXX CPU in this. We can see that the 4 core FX wasn't quite enough for this game, and the 8 core did fine, but what about the 6 core?

 

[slomo4sho]

I'd guess that was done to show frequency scaling on four cores, and to compare it to the slower hexacore with the same architecture.

The 4300 or the 6300 might not have been tested because Don didn't have them on hand.

Unless you're looking for conspiracy...

Just pointing out dependencies. I am also surprised that they failed to include any Haswell or Ivy benchmarks but resorted to using Sandy Bridge...

@slomo4sho: Mantle should work with all GCN cards.

Mantle is still unreleased and no real performance information is available... I see no practical reason to examine this API when no specifications are available...

 

[logainofhades]

another game my 8800gts 320mb cant run

Spend $100 and pick up a 7770. Or go on ebay and find a used 5850/5870 for cheap. 5850 is between a 7770 and 7790 and a 5870 is close to 7850 speeds.

 

[logainofhades]

The THG results seem to differ from some of the benches previously seen for some reason. http://www.tomshardware.com/answers/id-1823222/battlefield-beta-4670k-amd-8350.html