Another speedup: now ~70fps@1920x1024

Seems working on the algorithm pays off. Todays version reaches an even higher performance.
Also better performance for the forest scene as well (960x512@~100fps)

OpenCL Happy Buddha Raycasting 1920x1024@50fps

Here a first glimpse of  a new optimization. The good old Happy Buddha scene (1024^3) raycasted at 50 fps at almost HD resolution (1920x1024). Video below. For 960x512 it renders at 150 fps.

Instancing: Combined AABB + Voxel raycasting

Here a first combined test by raycasting 10.000 to 100.000 models.

Due to the mixed kernel, performance is not as good as pure octree.

Blending two Voxel Scenes

Here a experiment to apply blendshapes to a voxel scene. Its much slower, still runs at interactive framerates (20-30). The video lags as the framecapture@1024x768 took quite a lot of CPU time.

OpenCL Voxel Octree Raycasting

Here the same scene visualized using OpenCL octree raycasting. Eventhough the persistent threads method is not used, its around 20-100% faster than the previous splatting (depending on the scene). The code is ported from my previous CUDA raycaster. Hardware is again the GTX580M.

Each method has its advantages though. While octree raycasting works better for static objects, splatting is better for dynamic scenes.

OpenCL Voxel Splatting 2.000.000.000 points/s

This time OpenCL is used for rendering the voxels as single points. The gaps between are filled in a post-process. The performance is much better than OpenGL: 2 Billion Points/sec @ GTX580M

Resolution:1024x768 Voxel Field: 20.000 x 20.000 x 1024 Voxels, Framerate: 30-40 Fps, LOD and Frustrum Culling are used. Dataset: 5.4M points  / instance at highest resolution , 6.9M including all levels of detail.

Download : Link1 Link2

Voxel Splatting using OpenGL & Shaders

 

This screenshot shows a first test of using OpenGL 

and the Geometry Shader as well as LOD to render 

a voxel scene. 

Specs:

* 4 bytes / voxel 

* ~40 fps for the scene on the left

* LOD support

* Per voxel frustrum culling

Main bottleneck seems to be the triangle setup after the Geometry shader. As for the memory consumption, 

4 bytes / voxel is more than RLE ( 3.5 ) and less than the previous posted octree method ( 6.8 )