Except that's not what the architects had said, they specifically refers to the CUs in the context of graphics. Decoding is gpgpu ops which scale with the CUs, just because the numbers look aligned doesn't imply that they are related.
the bottleneck of the system is always the slowest part, hence the name, "bottleneck," in this particular case, the Juguar CPU is the weakest link, so it doesn't matter if you have more GPU power, the whole system can only compute as fast (or slow) as the CPU can.
So, you are actually agreeing with me XD
Their claim was:
"Everybody knows from the internet that going to 14 CUs should have given us almost 17 per cent more performance, but in terms of actual measured games - what actually, ultimately counts - is that it was a better engineering decision to raise the clock. There are various bottlenecks you have in the pipeline that can cause you not to get the performance you want if your design is out of balance."
"Interestingly, the biggest source of your frame-rate drops actually comes from the CPU, not the GPU, Adding the margin on the CPU... we actually had titles that were losing frames largely because they were CPU-bound in terms of their core threads. In providing what looks like a very little boost, it's actually a very significant win for us in making sure that we get the steady frame-rates on our console."
I said, contrary to what DrJay24 commented, that this particular benchmark falls back in what Ms claimed. And judging by what you said now, you actually agree, perhaps you though I was using this example to over generalize or something like that, sorry if it was the case.
The only thing it cements is that in one specific implementation of a video codec / player that most likely isn't even taking advantage of the extra cu it doesn't matter.
It couldn't be farther from cementing anything.
My comment was specifically in the context of DrJay24's quote, and we both were talking about specifically about this video codec.
He said that this shows how Ms was wrong and more CUs make for a better gpu, and I said that this bench doesn't show that because part of their claim was that everything has to be balanced, and this falls back to it.
Must be why AMD makes 30CU cards. Of course it scales, just like CUDA cores scale. How they scale is the only real question. 12 in no magic number, unless you consider die space, heat and cost.
Who's making that claim that you can never get more performance if you go beyond 12 CUs? Of course you'll have a performance increase, the point is that you need to scale other parts of your design too, being too much ALU and then being bound by something else that is far slower is not an ideal scenario.
Putting 30 or even more CUs on Xbone and Ps4 wouldn't increase performance linearly with the CU count, it might help, but there are more effective ways to get performance up.