Do you own an AMD processor and play Battlefield 4 64 player?

[cbn]

Fx in bf4 does well. Problem is on dx there is still that dreaded driver thread that cogs the thread rather easily on 64p mp and incurs in a lower min fps perf than your normal i5 or i7 solution.

This is because the AMD graphics driver is single threaded?

Yeah, I was just thinking what does a person use with AM3+? Nvidia because of the multi-threaded graphics driver (but can't use Mantle) or AMD (single threaded driver) but allows Mantle?

 

[PPB]

No, it is something inherent to dx11. Dx12 fixes this

 

[SPBHM]

it's not very simple,

with a lot of CPU power NV DX11 can beat even mantle

with a "slow" CPU mantle seems to be king

same for the FX 8 core but with a smaller margin.

 

[cbn]

No, it is something inherent to dx11. Dx12 fixes this

I brought that up because I noticed in your original post you mentioned Nvidia having higher minimum FPS under DX11, but it does make sense to me that with a multi-core processor (with relatively weak single thread) Mantle would win overall.

 

[PPB]

Pclab is a joke of a review site, no use bringing them to this discussion.

 

[SPBHM]

Pclab is a joke of a review site, no use bringing them to this discussion.

joke was testing Mantle without MP or enough different CPUs and Nvidia GPUs, like many websites did...

unless I can find other sources that contradicts them clearly I still rate them a good source for tests others are not really doing.

 

[PPB]

Sorry, i like my mp 64p servers full rather than empty, so they are as clueless as the ones doing sp mantle tests.

 

[Stuka87]

I have built a new PC now, but played BF4 64 player games for well over a year.

I had an AMD 965BE running at 4GHz, 12GB of RAM, an OC'ed HD7950 and played at 1080P.

Playing with DirectX sucked, many maps would drop to below the 30fps area. The GPU would be running at 50% load, the CPU just could not keep up. It would be running at an average of 95% load.

Playing with Mantle, things got *WAY* better. The game was much smoother, and stayed close to 60fps most of the time. My average FPS was in the 55fps range.

Of course now I have the same GPU, but with a 4690K at 4.5GHz, and I average 85fps at the same settings in Mantle mode.

 

[Stuka87]

This is because the AMD graphics driver is single threaded?

Yeah, I was just thinking what does a person use with AM3+? Nvidia because of the multi-threaded graphics driver (but can't use Mantle) or AMD (single threaded driver) but allows Mantle?

Last I checked almost no games use multi-threading for the driver because of the issues it causes. FarCry 3 did at first, but a later patch removed it. BF4 does not use it.

 

[SPBHM]

Sorry, i like my mp 64p servers full rather than empty, so they are as clueless as the ones doing sp mantle tests.

I didn't find any references in their test page to using empty servers, they only say they are testing in Siege of Shanghai MP 64!?

Of course now I have the same GPU, but with a 4690K at 4.5GHz, and I average 85fps at the same settings in Mantle mode.

DX and Mantle should perform a lot closer with this CPU

 

[lixlax]

I have:
860k
R9 270 OC 2GB
8GB Crucial 1600MHZ, runs at 1333MHz though because 1600 crashed a lot when playing the game.

I have some serious perfomance troubles after the last patch with the fps dropping to 5-20 freuqently, I havent figured out the cause.