Skip to content

Instantly share code, notes, and snippets.

@mikadou

mikadou/PerformanceBenchmark_AMD.md

Last active January 6, 2016 17:00
Show Gist options
  • Save mikadou/5c956ddedc56d485cb63 to your computer and use it in GitHub Desktop.
Save mikadou/5c956ddedc56d485cb63 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
PerformanceBenchmark_AMD.md

System specification:

  • Windows 7 64 Bit
  • Processor AMD Phenom(tm) II X4 945 Processor, 3000 Mhz, 4 Core(s), 4 Logical Processor(s)
  • 4GB RAM

Tested in Release mode:

Multithreading Benchmark RoR Version FPS
yes 3 Gavril MV4 0.4.5.1 10
yes 3 Gavril MV4 latest (b7b9810) 15
yes 6 Gavril MV4 0.4.5.1 5
yes 6 Gavril MV4 latest (b7b9810) 8
no 3 Gavril MV4 0.4.5.1 5
no 3 Gavril MV4 latest (b7b9810) 8
no 6 Gavril MV4 0.4.5.1 2
no 6 Gavril MV4 latest (b7b9810) 3
@Hiradur
Copy link

Hiradur commented Jan 2, 2016

What do you get in 0.38?

@mikadou
Copy link
Author

mikadou commented Jan 5, 2016

Could not load the exact benchmark in Version 0.38 because terrn2 is not supported.
I used SimpleTerrain and manually spawned 3 Gavril MV4.

Used prebuilt executable of Version 0.38.67.

Benchmark RoR Version FPS
3 Gavril MV4 (manually placed) 0.38.67 4-5

@Hiradur
Copy link

Hiradur commented Jan 5, 2016

Do the mv4s in your selfmade test touch/collide so they trigger the collision code?

@mikadou
Copy link
Author

mikadou commented Jan 6, 2016

Yes they definetly were all colliding. The low framerate made it very difficult to accurately place the vehicles. Therefore the setup was much more chaotic than the original benchmark, but collision between cars was given.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment