Previous threads for reference:

https://www.reddit.com/r/overclocking/comments/1ot2y6k/9800x3d_vs_14900ks_benchmarks_in_high_cpuusage/

https://www.reddit.com/r/overclocking/comments/1oydb2f/updated_9800x3d_vs_14900ks_benchmarks_in_high/

Recap: With many recent games with RT being CPU-heavy, like Outer Worlds 2, I decided to test out 14900KS vs 9800X3D in 5 4* RT/PT games that I had played. With CapFrameX, I took 3 one minute runs in CPU-limited areas and then aggregated them.

*Removed Stalker 2 results since it left GamePass. It also had a big performance patch around the same time and the older results will not be useful.

All RT/PT settings were enabled at 4k 1440p w/ DLSS Ultra Performance on a 5090 with the latest DLSS4.5 driver.

Results : 14900KS results were unchanged despite the resolution and DLSS preset change, except for Oblivion, and I simply reused the older results.

9800X3D saw 3-5% improvements with 8000 RAM, except in Oblivion where it is within margin of error. This is likely due to improved latency since the Infinity Fabric limits bandwidth to same as it was on 6000 RAM.

9800X3D is now definitely faster in averages, though 1% still favors 14900KS. Stalker was the 2nd worst result for 9800X3D before, so its removal also helps.

Details :

9800X3D w/ 8000C36-48-48-60 2x24GB on Apex X870E ( FCLK 2000 )

By default SMT is enabled

PBO set to auto

14900KS w/ 8400MT/s C38-49-49-69 2x24GB on Asrock Z790I Lightning

HT off, only 8E cores enabled

59/42/50 for P-cores, E-cores and Ring respectively

AC/DC loadlines set to 0.40/0.74