HIGHLIGHTS
SUMMARY
The authors provide an in-depth study of the anatomy of six state-of-the-art APIs for GPGPU programming: OpenGL Compute Shaders, Direct Compute over Direct3D11, OpenCL, CUDA and next-generation bare-metal APIs Vulkan and Direct3D12. Out of this study the authors derive a taxonomy of the semantics supported across the considered APIs, highlighting how the low-level services underlying virtually every available API for GPU development largely overlap in functionality. The authors also include numbers from a seventh API, the CUDA runtime API, which is not wrapped by the implementation of . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.