Does CUDA support 16-bit (half) floats?
All floating point computation is performed with 32 or 64 bits. The driver API supports textures that contain 16-bit floats through the CU_AD_FORMAT_HALF array format. The values are automatically promoted to 32-bit during the texture read. 16-bit float textures are planned for a future release of CUDART. Other support for 16-bit floats, such as enabling kernels to convert between 16- and 32-bit floats (to read/write float16 while processing float32), is also planned for a future release.
Related Questions
- The Play England regional offices will be established over the latter half of 2006. How will Play England provide regional support the local authorities who are applying in July and November 2006?
- Does SheerVideo support film, 2K, Cineon Half, or Cineon resolution?
- Does CUDA support 16-bit (half) floats?
