The test took place in the games Alan Wake 2, Total War Warhammer, Call of Duty: Warzone 2, CS: GO 2, Cyberpunk 2077, Fortnite, Hogwarts Legacy, Starfield and The Last of Us: Part 1. There was also Cities Skylines 2, there but strange results were obtained, so I preferred to skip this test (there were some strange results elsewhere, but they were not so extreme). And how did it actually turn out in this set of games?
Of course, it depends on the resolution, and in the area of the average frame rate, the novelty is only 1% slower in 4K. At 1440p the difference is less than 5% and at 1080p the difference is over 7%. In some games, the new processor is in the lead, but in some it is even more than 15% behind, in any case, the average differences here are between 1 and 7%.
But if we look at the 1% FPS metric (the worst 1% of results), the power of 3D V-Cache starts to show. Although this does not cause a significant increase in the average frame rate, it significantly helps to keep the lowest values achieved at higher values. Here, the novelty already loses between approx. 6 and 14%.
If we look even further at 0.1% FPS, the situation is quite similar. The 1080p results in Starfield are strange here, where I suspect a resolution flip in the graphics. We get a 7 to 14% difference here. If we assume an error in the graph and do not use Starfield, the differences will be 9 to 17%.
As we can see, the 9900X can’t overcome the 3D V-Cache advantage of the previous generation of processors according to these first tests in games. In the case of average FPS, the difference is small, however, in the metric of 1% and 0.1% FPS, the difference is more pronounced, and here the advantage of a larger cache is more obvious. However, I would still take this test with a grain of salt and wait for some more extensive reviews, which will hopefully not have somewhat surprising results and will be in a larger number of games.
Source: www.svethardware.cz
