Beam Performance with Multiple Core configurations on 285K *Interesting results*

As I continue to keep testing out Beam and its performance on different GPU/CPU's I thought I'd try be changing the core counts on my 285K The thought that Beam always wanted more cores still plays true but the game engine does some interesting things.

Setup is as follows

Core Ultra 9 285K 8 P-Core 16 E-Core (No OC)
MSI Pro B860M M/B
32GB DDR5 7200 T-Force Delta Ram
Asus TUF RTX 5090 OC
2TB Samsung 9100 Pro SSD

I ran all tests in both direct 3D and Vulkan Mode

I took frame rate data at different points in the following maps:

West Coast
Automation
City of Los Injurus



What's most interesting is that the 1P 16E configuration actually gives the best frame rates when in Direct 3D mode. Once you change to vulkan mode it's faster in half the situations tested. It's also becoming apparent that Direct 3D mode is probably going to be close to a dead end because when using high end hardware Vulkan takes advantage of it so much better.

What I find really interesting and goes against what my other testing has shown is the performance with 8 Performance cores is actually not very good.

I'm able to see how the performance of beam scales based on clock speed (using my Core Ultra 225 as an example) an extra 700mhz on the 285K P Cores was good for 15 fps in west Coast vs the 225.

Beam uses the Performance cores only until the spawned cars approach the physical amount of the P-Cores then transitions to the E-Cores, so spawning a single car I can see how clock speed affects the performance (using the same cpu architecture of course). I guess the performance is impressive considering I can get playable frame rates with a bunch of cars out of only 8 cpu cores, i just didn't think I'd get a performance regression.

What's really promising is how the performance scales across the 16 E-Cores, good job devs on making this game take advantage of high end hardware.