Analysis of DX12 thread usage in the Rise of the Tomb Raider DX thread discussion
Analysis of DX12 thread usage in the Rise of the Tomb Raider DX thread discussion
I rechecked the performance at 1080p using identical procedures but at 3.7Ghz to allow dual-core and multi-threaded operation. I applied both minimal settings with maximum detail, which demands more CPU power, then performed additional tests with all parameters increased except textures, which were set to high quality, no motion blur, and anti-aliasing enabled. I omitted the 3-core test after observing a consistent pattern. Beginning at 4 cores and reducing, some games failed to load at all levels, leading to higher frame rates as expected. This likely stems from the game's DX12 implementation or the benchmark design itself. It’s important to remember that models missing in lower core counts were an issue in earlier tests; just because FPS numbers appear doesn’t mean they’re accurate. Avoid using a dual-core setup with Tomb Raider. Quad cores must be properly tuned via settings to fully utilize GPU resources.
Generally accurate, though I noticed more glitchy models in 4-core 4-thread configurations compared to 4-core 8-thread ones. Also observed 6-core 6-thread outperformed 4-core 8-thread, a detail the graphs don't fully capture. However, at higher resolution tests, a 1:1 core-to-thread ratio is preferred, even though adding SMT to the four or six-core set doesn't seem to cause major issues.
This particular game doesn’t offer a strong incentive to skip using dx12. It might actually perform better on a dual-core setup.