Epilogue
This review took me almost a week of testing because I conducted several game benchmarks repeatedly to ensure everything was in order. Moreover, it took me quite some time to gather and analyze all the data, although I have prepared some special Python scripts for this purpose. Again, I wanted to be extra sure of all the results, so I did lots of manual checking.
| Game Title | Raster – HD/QHD/UHD/Average (24H2 improvement – %) |
Ray Tracing – HD/QHD/UHD (24H2 improvement – %) |
Ray Tracing – DLSS/FSR – HD/QHD/UHD (24H2 improvement – %) |
Ray Tracing – DLSS/FSR – Frame Generation – HD/QHD/UHD (24H2 improvement – %) |
|
| Alan Wake II | -3.28/-2.44/-2.23/-2.65 | -0.74/0.45/-0.32/-0.20 | -4.24/9.47/20.41/8.55 | 5.54/8.67/-0.74/4.49 | |
| Assassin’s Creed Mirage | -6.14/-4.2/-2.0/-4.11 | ||||
| Avatar: Frontiers of Pandora | 5.65/4.24/1.1/3.66 | 11.81/7.93/5.64/8.46 | -23.38/-20.0/5.65/-12.58 | ||
| Cyberpunk 2077 | -3.53/-2.16/-2.12/-2.52 | -1.95/-1.68/-2.48/-2.04 | -2.94/-0.63/-0.25/-1.27 | -4.43/0.24/-1.95/-2.05 | |
| Dead Space | -5.73/-0.57/-0.47/-2.26 | -10.66/-2.22/-0.45/-4.44 | -8.12/-8.8/-0.61/-5.84 | ||
| Forza Horizon 5 | 0.61/2.38/0.86/1.28 | ||||
| God of War | -3.19/-2.63/-3.44/-3.09 | ||||
| Hogwarts Legacy | -6.50/1.21/1.40/-1.30 | -11.73/-2.9/-3.09/-5.91 | -11.9/-16.78/-3.38/-10.69 | -6.17/2.34/-10.66/-4.83 | |
| Ratchet and Clank | -3.13/-1.33/-4.48/-2.98 | -13.75/-5.16/7.71/-5.43 | -17.04/-9.13/1.19/-8.33 | 11.40/-1.61/0.61/3.47 | |
| Resident Evil 4 RE | -7.72/-1.82/-1.28/-3.61 | -7.99/-1.78/-0.12/-3.30 | -7.89/-7.74/-2.37/-6.0 | ||
| Starfield | -1.29/-2.58/-2.01/-1.96 | ||||
| The Last of Us Part I | -3.74/-1.39/-1.3/-2.14 | ||||
| Average Improvement (%) | -2.662 | -2.0475 | -2.16 | -2.3 | |
| Total Improvement (%) | -2.29 | ||||
As you can see in the table above, the total improvement that the 24H2 offers once paired with an AMD processor is negative, meaning that you don’t get any actual performance upgrade! So I can’t help but wonder why AMD states the opposite! Of course, the performance difference between the 23H2 and 24H2 has to do with the games that you will select, but it doesn’t make much sense to go with older and less popular titles. The games I used for comparison purposes cover almost all popular game engines, providing a vast insight into what is going on with AMD statements about a notable performance improvement with the 24H2 Windows update. Not only did I not see this update, but I also noticed an overall performance drop.
You might notice a performance improvement in some specific titles and under conditions related to the game settings, but this is not the case for most titles I tried. So, whoever rushed to back up AMD’s claims for a significant performance increase with this Windows update should check twice and not rush to post results that confuse users.
In my 7800X3D re-visit review, a viewer commented that my results were outdated and that I should check with the 24H2 update, although it is NOT available yet. I know that on the Internet, everyone has an opinion, especially when he writes anonymously, but I can’t help it and get pissed when I read such stupid and unfair comments about my work. This pushed me to proceed with today’s article and offer you a complete insight into what you should expect from the 24H2 update.
Look at our YouTube channel for the video version of this article!
as a decade old pc enthusiast just coming back I can’t help but see these benchmarks at 300fps and wonder what are ppl smoking? no way…but throw unique titles like tarkov in here and you’ll see it drop below 100fps even on your x3d chip, now this is a game that people reported gains in fps/consistency when installing the Ryzen KB update on 23h2 which is built in 24h2
These tests make no sense. You’ve invalidated them by not using like-for-like drivers. You’re not so much comparing windows as you are comparing driver configurations
No sense to you probably.
The performance difference between driver versions can be upwards of 10%. This test is worthless.
Show me your results proving that! That the OVERALL difference can be 10% between 3-4 month driver versions! Your comment is totally useless without proof!
i think your test have problem
you test on old , modify 23h2 not new defult windows iso ( you change many settings before )
you must instal new 23h2 publish 1 year ago , rathen test and will see 20% performance on 24h2
other thing , you must run game on 1080 medium quality (cpu heavy) not ultra max settings (gpu heavy)
i like see this test on AMD GRAPHIC CARD , i am sure when amd cpu ,gpu give more performance / price
Haha wait your saying they should test an un optimised install from a year ago? Which all gamers would have installed not tweaked to the best it can be?
that doesn’t make any sense.!