Going off topic now so i will rein it in after this before i get whacked by a mod.
This will surprise people but the 2013 Xbox CPU, with the 8 Jaguar cores at 1.75ghz had the approximate CPU compute power of an K1 Nvidia Shield or, a 2012 Google Chromebook.
In short, the 2013 Xbox Cpu was awful at launch. the IPC, Clock, Architecture, ram bus speed - the lot.
In CPU terms, this new 2020 Xbox will be as fast as the same 2020 Xbox running a 3900x at the 99th percentile, in the majority of gaming scenarios.
We have proven it on the forum time and time again in our own independent performance testing, games programmers don’t (with small exceptions) use all the extra cores they are given beyond 6. Games are not as multithreaded as they should be.
Our £7k rigs worked well in 3dmark and Cinebench but they meant little when running Batman or, Crysis benchmarks, IPC mattered, not cores.
If you took a 3900x and plugged it into the socket into the 2020 Xbox X and did a blind test against the standard version, on a TV, in-game, I doubt anyone would notice the difference.
Hence, my claim that we have console gaming performance parity with a Desktop PC is simply because the IPC performance of the Zen2 heart and the clock speed is now as fast as a desktop for its intended purpose - playing games.
Microsoft and Sony could have gone quicker on the CPU clocks but i bet that in their internal CPU clock testing, they came to the same conclusion that this board came to years ago - CPU clock only gets you to a certain FPS point and beyond that, you are burning heat budget to beat a benchmark, the actual user experience is not improved and you are better off either upgrading your CPU architecture or, more effectively, spending the heat budget on GPU clock and faster, hotter, VRAM.
