Get the latest Blender, older versions, or experimental builds.
Stay up-to-date with the new features in the latest Blender releases.
Access production assets and knowledge from the open movies.
Documentation on the usage and features in Blender.
Latest development updates, by Blender developers.
Guidelines, release notes and development docs.
A platform to collect and share results of the Blender Benchmark.
The yearly event that brings the community together.
Support core development with a monthly contribution.
Perform a single donation with more payment options available.
Reorganize code to reduce register pressure.
While cubic interpolation is quite expensive on the CPU compared to linear interpolation, the difference on the GPU is quite small.
* Use common TextureInfo struct for all devices, except CUDA fermi. * Move image sampling code to kernels/*/kernel_*_image.h files. * Use arrays for data textures on Fermi too, so device_vector<Struct> works.
