It's funny, you people are toting more cores and threads on AMD like they are god's gift to the computer world. Currently yes, AMD wins in performance per dollar and they do have more cores.... But the even the fact that Intels CURRENT cpus still manage to beat out AMD's BRAND NEW cpus in gaming is kind of embarrassing in my opinion.
Honestly, what do you people do that you need "MOAR CORES!"? 'Cause last I checked the performance gain from "MOAR CORES!" really isn't all that great. Especially when you compare AMD and Intel on a core vs core basis.
You are touching an important subject here. More cores is
better, and we do need more cores. Still, faster* cores remains even more important. Even i7-7700K beats Ryzen 7 1800X i more than gaming, even in major use cases like Photoshop and web browsing. And an i7-6800K would beat 1800X in most normal power user applications, with a few exceptions.
*) So why are Intel's cores much faster when Ryzen have more computational resources in each core? Simply because Intel have much a better prefetcher to feed their CPUs. In the real world this means Ryzen will handle simple linear streams of data with no branching (software rendering, compression, some encoding), while Intel remains superior on "mixed" workloads which are not cache-optimized (gaming, photo editing, CAD, web browsing, office applications etc.). Most use cases for power uses resides in the latter end of that scale.
Most people have misconceptions about how games/rendering engines work. Even though rendering is a parallel task, dispatching the queue is not. Having multiple cores building a queue is possible in Direct3D 12, but the synchronization overhead is going to outweigh any gains, so you are pretty much limited to one core per queue. Multithreading in games is primarily about letting the rendering work undisturbed, not about splitting the rendering itself. So gaming performance are still going to come down to having cores that are fast enough to not be a bottleneck for the GPU, but no scaling beyond that point. Current Intel CPUs hit that sweetspot around ~4 GHz (boost), which is why CPUs like i5-7600K, i7-7700K, i7-6800K and i7-6900K all perform "the same" in gaming. But you don't see this with Ryzen, even >4 GHz they are still not fast enough, even with plenty of cores and computational power. This is, as mentioned, caused by the prefetcher. Most games' rendering code is suffering from branching and cache misses, both of which causes stalls for the CPU. Having a better prefetcher helps mitigate the performance penalties from this, of course only to some extent. Going beyond 4 GHz would not make up for this for Ryzen, since the penalties for cache misses are time constant.
