PS6 performance gains may have been overhyped

[Redaktion]

When it comes to performance uplift for the next generation of consoles, leakers can't seem to agree. Known leaker and insider KeplerL2 has called out another leaker and insider, MLID, for misinterpreting AMD's internal performance numbers.

https://www.notebookcheck.net/PS6-performance-gains-may-have-been-overhyped.1276883.0.html

[5x path-tracing or gtfo]

Only path-tracing, not ray-tracing, is visually transformative tho.

(Full node jumps: 7nm -> 5nm -> 3nm -> 1.4nm -> ..)
Expecting 2 full node jumps from the PS5 (TSMC 7nm node variation) to PS6 (TSMC 3nm node variation) is reasonable (but it would also be only 1 full node jump from the PS5 Pro to PS6). Each full node jump gives about 25-33% better energy efficiency or FPS per Watt. So you go from e.g. 60 raw (=no up-scaling) FPS to 60*1.33*1.33 = 106 FPS.

[Jay Mann]

This article is filled with typos. That aside, I think this is weird, the PS5 Pro is already advertised as having 3x Ray Tracing performance and since the original PS5 RT performance was low since AMD's RT technology was immature, is not illogical that the PS6 could 10x the RT performance of the original PS5. That being said 10x Ray Tracing performance does not mean 10x frames per second I don't know who thought that but as the article states the leaker never claimed that. It seems people misunderstand what 10x performance in a single specific part of the rendering pipeline entails and just assumed 10 times the fps