From what I understand, it's more accurate to say that it reduces CPU use by about 50% (rather than improving performance). If CPU is not your bottleneck, you won't see a performance improvement. If it is (which is common these days), then you'll have twice as many free CPU cycles on your main thread, which will translate into however much improvement your GPU can muster.

Even if CPU isn't the bottleneck, you ought to see some power savings.

Either way, it will make it easier to bolt a monster GPU onto a smaller, more efficient CPU.

(What we do about GPU power consumption is a different problem...)

> From what I understand, it's more accurate to say that it reduces CPU use by about 50% (rather than improving performance)

According to this blog entry, you are mistaken, and it really can improve performanec 54% in a multi-GPU rig.

http://battlelog.battlefield.com/bf4/news/view/bf4-mantle-li...

There seem to be issues with FPS drop, though ('stutter').

The 10x improvement that Apple claims is presumably rooted in Mantle's design. The Mantle docs also claim an order of magnitude improvement in draw calls.

See the "Small Batch Problem" section in this PDF: http://www.amd.com/Documents/Mantle_White_Paper.pdf

Sadly BF4 is kind of an optimal case for an API like Mantle. Their rendering workload is heavily CPU-bound, so Mantle produces big gains. A lot of other game engines will see much smaller improvements because all the weight is on the GPU, or on systems Mantle doesn't touch, like physics.

Any renderer which isn't CPU (read: driver) bound doesn't take performance seriously or hasn't been optimized yet.

(Not that that refutes what you're saying, I'm just saying that most serious engines are in the same category as BF4 when it comes to gains from better driver APIs)

I'm less interested in performance boosts than in improved driver stability/consistency from a simpler driver model.