Skip to content

ChrisMisan/IFS-project

Repository files navigation

IFS-project

IFS fractal explorer

!!!ATTENTION!!!
IT IS STRONGLY ADVICED TO MERGE TDR WITH REGISTER BEFORE USING CUDA ACCELERATED RENDERING, GPUs are loop based and drivers will automaticallly kill any process which uses gpu for more than several seconds, we really don't want that in general computation solutions.

Description

Compressed IFS fractal explorer which uses multithreading or GPU CUDA for calculations. Rendering pipeline implemented via SFML lib with few OpenGL tweaks, UI implemented using nuklear lib with SFML as rendering and baking engine.

Keep in mind that GPU rendering is only avalible for NVIDIA GTX/RTX series and won't work with any AMD GPU. High quality render is avalible only in 2k resolution, it is advised to use resonable samples amount, in most cases 10 billions should be more than enough.

Expected render times for HQ:

Keep in mind it was measured some time ago.

  • GPU-under 5 min
  • Multithreading (depends on physical core number and speed), for 4 cores around 20-30 min
  • Single core (depends), for 3.2 GHz around 1.5h

REMEMBER: Multithreading and GPU rendering affect only HQ, preview mode will bypass any multithreading.

For now explorer supports only 8 bit grayscale but renders could be easily colored using postprocess softs thanks to 3D nature of the algorithm.

HQ render example

ScreenShot

UI screen

ScreenShot

Implementation

Languages: C, C++
Dependencies: CUDA, SFML, nuklear
Owner: Christopher Misan

About

IFS fractal explorer

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published