I'll start I was surprised to see OpenGL vs Vulkan results. While OpenGL had a better score and much higher max. FPS, it also had much lower min. FPS, which was unexpected. I'd rather have a stable framerate than a higher max, so Vulkan is the winner for me.

Specs:


OpenGL:


Vulkan:


Overclock:
CPU: 3.3GHz -> 4.4GHz
GPU: +125MHz graphics clock, +300MHz memory clock

Info:



OpenGL:


Vulkan:


4K is bit hard for GPU like this. 1080p or 1440p is more suitable on GPU's below GTX 1080 or Vega 64


Ryzen 7 1700X @3.8ghz
GTX 1070 396.24.02 drivers, core offset +100, mem offset +300


just for fun a more suitable for this GPU - 2560x1440 result:

Xpander4K is bit hard for GPU like this. 1080p or 1440p is more suitable on GPU's below GTX 1080 or Vega 64

OC: GPU: +150MHz graphics clock, +150MHz memory clock

Specs:



OpenGL:



Vulkan:

Update did not have powerlimit increased from 150W to 170W on previous test, so the clocks did not go that high in reality.

same +100 core and +300 memory

OpenGL:


Vulkan:


now Vulkan minimum seems to be better also. Also i noticed that Vulkan rendering doesnt use CPU allmost at all, just 0.9-1.5% while with openGL i see over 6% of usage (thats not htop numbers but from 100% with 16 thread)

Not sure why basemark is not able to detect my Arch installation.

However I'm really surprised about the notable difference between OpenGL and Vulcan.

You can see my system in my PC Info profile. Everything is running stock. Only exception are the memory modules with does NOT run with their performance profile and clocked defensively to provide maximum stability. I'm not into OC at all.

SPEC


OpenGL


Vulcan

Wow, we have a new #1! The difference between GTX 1080Ti and RTX 2080 seems crazy.

The Vulkan one did not run for me. I'm not even sure that the OpenGL one ran correctly. Instead of a large graphic animation, it was 64 square tiles of changing images.

14Instead of a large graphic animation, it was 64 square tiles of changing images.

Vulkan:



OpenGL:



Info


GPU in stock settings.