Search for:
Benchmarking DX11 vs Vulkan Performance: October 23, 2025 Update
CS:GO

Benchmarking DX11 vs Vulkan Performance: October 23, 2025 Update

By


Previous update's results can be seen here

Benchmark template spreadsheet

Please put (or edit) your results in this form

Purpose of Benchmarking

To my understanding, Valve intends to eventually drop DX11 support in one of their in-development games: Deadlock (source). Because Deadlock doesn’t have a benchmark map or program, I decided that the next best option would be to benchmark CS2 in both DX11 and Vulkan every time an update has anything related to performance improvements in the changelogs. The goal is to document if any improvements or regressions have occurred in Vulkan or DX11 performance in CS2, and to see if this could indicate any progress towards a full transition to Vulkan in later Source 2 games. A link to the template spreadsheet can be seen above, and my benchmarking results will be provided below along with a commentary.

My Non-Game Specs

  • CPU: AMD Ryzen 7 9700x
  • GPU: NVidia Geforce RTX 5070 Ti
  • Operating System: Windows 11
  • GPU Driver Version: 581.57

Benchmark Results

Link: https://docs.google.com/spreadsheets/d/1yNiZusJyI4MNtRq2gPksICkD_UaxN3GEXLJBxpBojX8/

  • Mean DX11 Average FPS: 696.83 (Standard Deviation: 8.92)
  • Mean DX11 1% FPS: 198.5 (Standard Deviation: 5.85)
  • Mean Vulkan Average FPS: 671.07 (Standard Deviation: 7.48)
  • Mean Vulkan 1% FPS: 189.3 (Standard Deviation: 3.33)

Discussion

Since the previous post, I've added a standard deviation row to the spreadsheet to compare the stability of the two renderers. I will also add this to the google form, so if you've answered it already, make sure to update. (For my system at least), the raw numbers of both DX11 and Vulkan have gone down. However, with the October 22 and 23 updates' "stability improvements", variance between each benchmark's results have overall gone down. DX11 became somewhat more consistent in its 1% lows, while Vulkan's 1% lows are now incredibly stable (at the cost of average fps being almost as all over the place as DX11 is for me, although still lower). Since I believe that Valve cares more about a smooth and stable experience than one that can occasionally touch a high FPS, I think that they'd see this as a partial win for Vulkan, as the 1% lows being so consistent means that the player should have a consistent experience. However, given that Vulkan is supposed to offload the CPU load, I hope to see Vulkan's numbers climb higher in my cpu-bound results (I am essentially running the game in 360p) while still having the same level of consistency as seen right now. I'm aware that I am an outlier, having a recent mid-high end setup, so I would appreciate more people testing and documenting any changes they have here.