Hardstuck-Platinum said:
No it's definitely CPU related. As the GPU in the XBSX is 3x better than the one in the XBSS, any game that's GPU limited to 30 on the S would hit 60 no problem on the X. Problem is though, they have the same CPU so any game that's CPU limited to 30 on the S will also be CPU limited to 30 on the X. MS should have designed the X with more powerful CPU to avoid this 30FPS problem |
I agree this is most likely a CPU bottleneck but Xbox simply could not make the CPU stronger on Series X than it is.
First, because this would have created way more issues than anything else. Keeping the CPU very close ensures that games are the same feature-wise and that the divergence happens almost exclusively with the presentation. There is some leniency thought because when rendering scenes with lessened resolution/and simplified assets the CPU would be slightly less used (things like fewer draw calls would be expected) and it's actually what we see with series S vs X CPU where the X is about 8% faster.
Secondly, the CPU does offer the best of what AMD could offer in 2020 for an 8-core variant when designing a ~200W max system and avoiding chip yield issues. Going above it would have meant either:
a) going to 12/16 core.
This would only have given mixed results, especially with 3rd party as the game must be written to take advantage of extra cores, and not everything is as easy to parallelize. This would also have increased die size significantly, reduced yields, and increased power requirement leading to a costlier, louder, even bulkier system.
b) Increasing Frequency
There is some headroom but it would only have been marginal, best case scenario would be +~20%, and it would either lead to a significant power consumption increase leading to a louder and/or bulkier system or a more drastic chip binning approach leading to reduced yields and significantly increase cost.
