About Linus' Mantle BF4 benchmark procedure

[TechFan@ic]

Since when is it "not fair" to compare competing graphics cards ? or any computer part for that matter.

Mantle is about reducing API overhead so the best way to go about testing that is to keep it simple really.
One high end GPU, an R9 290 or a 280X for example and two different processors.
A 7990 can be added to emphasize the CPU bottleneck in multi-GPU configurations as well.

 

An FX 8350 with 4,6 & 8 cores enabled to showcase if there is any performance scaling with additional cores like AMD is claiming.
On the Intel side an i7 3930K, again core scaling with 2,4 & 6 cores.

[warmmilk]

The new Mantle patch will not affect NVIDIA performance in any way.

 

The only thing I can think of is that NVIDIA might have a driver revision ready to roll with some BF4 optimizations that might help to narrow the gap.

 

Anyway, sorry we didn't read more tweets, but I've definitely read your message now

 

you should also include minimum FPS in your results...  if you haven't done the vid yet, if you have, include it in all your other gpu testing.  IMHO min FPS is just as important as average...