| Vodacixi said: I have a question from those like me who don't have enough technical knowledge. |
The key is not the difference in loading speed. Loading Gigabytes of data really fast is only the first part of the problem (from a logical standpoint, it is the no-brainer part). If you are already satisfied with that first part, you are in trouble.
The other part is "what do I have to do with the data once I have unpacked it into memory" If you look at Cerny's talk, it incorporates a lot of individual steps which are mandatory for any console or pc. The data does not magically unpack and be ready for the gpu. Apparently Cerny developed the whole chain in hardware, and that is a lot of powerful hardware, to the extent to the equvalent power of several Zen2 cores. At this time, we do not know what the XSX does after loading and unpacking the data. If they don't have the hardware (and they can't have all of it, as some pieces like the cache scrubber are Sony exclusive hardware), they have to do it in software with the Zen2 cores. That has very obvious consequences. They can do all of it with the cpu, but it requires several cpu cores. While Sony's Zen2 cores are free to do whatever they want, probably half of the Zen2 of the XSX is out of the picture (and let's not forget the caches of the Zen2 get a kick in the butt, too). So there is a (serious) performance penalty in the cpu part every time hires assets are streamed. How that translates into gaming performance is everyone's guess (tiny hickups might appear occasionally?), only the games will tell. My guess is it should be noticable and was worth the effort, as Sony would not have invested so much time and money into this pure hardware solution.
Last edited by drkohler - on 21 March 2020
