DirectX 11 includes these new high quality texture compression formats. The DX SDK has samples with compute shader implementations of compressors and decompressors of the new formats.
Good technical info on the formats:
BC6H
http://msdn.microsoft.com/en-us/library/windows/desktop/hh308952%28v=vs.85%29.aspx
BC7
http://msdn.microsoft.com/en-us/library/windows/desktop/hh308953%28v=vs.85%29.aspx
More details are available in the OpenGL spec:
http://www.opengl.org/registry/specs/ARB/texture_compression_bptc.txt
This presentation also includes some information on the new formats which are basically new interpolation methods:
http://www.nvidia.com/content/nvision2008/tech_presentations/Game_Developer_Track/NVISION08-Direct3D_11_Overview.pdf
BC6H - floating point data
High Dynamic Range
6:1 Compression (16 bpc RGB)
BC7 - 8-bit fixed point data
LDR with alpha
3:1 compression for RGB or 4:1 for RGBA
Comparison of the formats
http://www.g-truc.net/post-0340.html
Thursday, August 11, 2011
DirectX 11 : Compute Shaders
Resources on learning about Compute Shaders
This is an important presentation on the performance characteristics of compute shaders and things to look out for. Compute shaders are not necessarily faster than PS and CS settings need to be tuned for optimal performance on different GPUs.
http://developer.amd.com/gpu_assets/DirectCompute%20Performance.zip
This page has some good links:
http://www.danielmoth.com/Blog/DirectCompute.aspx
Good info on Compute Shader thread specifics and how they relate to real hardware:
http://developer.amd.com/gpu_assets/Efficient%20Compute%20Shader%20Programming.pps
http://openvidia.sourceforge.net/index.php/DirectCompute
Compute Shader Debugging
http://cmpmedia.vo.llnwd.net/o1/vault/gdc10/slides/Aguaviva_Raul_OptimizingGraphicsApplicationsWithAMDGPUPerfStudio2.pdf
Nvidia Parallel Reductions with CUDA (Compute Shaders)
http://developer.download.nvidia.com/compute/cuda/1_1/Website/projects/reduction/doc/reduction.pdf
This is an important presentation on the performance characteristics of compute shaders and things to look out for. Compute shaders are not necessarily faster than PS and CS settings need to be tuned for optimal performance on different GPUs.
http://developer.amd.com/gpu_assets/DirectCompute%20Performance.zip
This page has some good links:
http://www.danielmoth.com/Blog/DirectCompute.aspx
Good info on Compute Shader thread specifics and how they relate to real hardware:
http://developer.amd.com/gpu_assets/Efficient%20Compute%20Shader%20Programming.pps
http://openvidia.sourceforge.net/index.php/DirectCompute
Compute Shader Debugging
http://cmpmedia.vo.llnwd.net/o1/vault/gdc10/slides/Aguaviva_Raul_OptimizingGraphicsApplicationsWithAMDGPUPerfStudio2.pdf
Nvidia Parallel Reductions with CUDA (Compute Shaders)
http://developer.download.nvidia.com/compute/cuda/1_1/Website/projects/reduction/doc/reduction.pdf
Subscribe to:
Posts (Atom)