Support NeoGAF

[Naminator]

Belies the reality?

Lol.

Guess what, Win7 with a normal cpu ("unoverclocked" is even a word? as if there wasn't people that doesn't overclock their cpus!) and dx11 is the most standard configuration of OS and graphical API used in games. Hell, lots of games still use dx9/10.

Is it REALLY that hard to understand the issue here?

The 780ti is being bottlenecked by the CPU and Windows 7.

The title of the thread is at it's best disingenuous.

If we all feel fine with posting benchmarks where one piece of hardware is bottlenecked, lets then take AMD's word for it and say that their new APU performs just as well as a Haswell i5!

Who cares that the Intel CPU is being bottleneck by that low-range GPU, it's a totally valid comparison right?

 

[riflen]

What is being hidden?

It's not that there's anything deliberately hidden. It takes little effort to be misleading without hiding information. I have a problem with the following:

The thread title. What's being compared in the benchmark are APIs. You'll notice that the R290x under DirectX 11.0 gets spanked harder than the 780Ti. The title is pure flamebait, an attempt to provoke shit-talking.

The opening post contents. "a direct fight between AMD and Nvidia". Mantle is intended to be an open API, cross-platform and multi-vendor. These are AMD's words. Again, this is another attempt by the thread creator to try and stir up shit. This really has nothing to do with GPUs, the frame-rate is improved because of superior utilisation of CPU resources.

Finally, the benchmark itself is a bit daft. They are comparing Mantle to DirectX 11.0. DirectX 11.2 has significant improvements in CPU utilisation over 11.0. Would anyone be impressed if Mantle was compared to DirectX 10 and came out ahead? I can't read the article, do they state why they compare against an API from 2009?

 

[MadJackChurchill]

Make this applicable to every game and you may win me back AMD. Until then, no chance. I still remember the crappy driver support, lack of features, and the fact that my factory OCed 4870 would cause massive screen flicker if i enabled the OC. Not switching back to AMD just to play BF4.

Preach.

I've had five AMD cards in my time, each one had SOME problem: whether that is heating, constant crashes, driver issues etc.

 

[benny_a]

I can't read the article, do they state why they compare against an API from 2009?

In the comments the staff says that Windows 7 DX11 is the default setup that people run, so they thought it was an interesting benchmark to run. (I think based on what they've written that they perceive this to be a common setup. The official recommended OS off the staff is also Windows 7 SP1.)

The OP is representative of the article and the comments written by the website staff.

The thread title is about how the Mantle API performs. Based on the thread title and the article I see it as an impressive achievement of Mantle that a more than 200 EUR cheaper graphics card using Mantle outperforms the one using DirectX in a what they consider a standard PC build.

 

[artist]

So, you are recommending everyone to get very expensive CPU and overclock it?

Benchmark in OP is very much valid to vast majority of players. Even more so to those who have less than high end 3xxx i7.

Yes, 4.6GHz K CPUs are common place. I just looked up on Steam hardware survey and it's a solid as ghst's logic!

I spend some minutes and read through the comments by the authors. Doesn't seem like the OP has misrepresented anything.

This is exactly how they argue in their article and their comment section.

This.

 

[slapnuts]

wow amazing boost

 

[Aversion]

I'm sure it is a little bullshit but there is no question we need this low level access on PC. My 770 should be beasting stuff even with high AA where it has trouble. I hope Micro is serious about DX optimizations.