Namespaces and Classes

• Cekirdekler
• • Hardware: namespace for explicitly selecting platforms and then devices (newly added with v1.1.x+)
• • Pipeline: namespace for managing device to device pipeline operations with ClPipelineStage and ClPipeline classes.
• • • SingleGPUPipeline: namespace for a single device pipeline builder with a more efficient work scheduler algoritm.
• • • Pool: namespace for building efficient non-separable kernel pipelines for multi-GPU systems.
• • ClArrays: namespace for creating C++ array wrappers that work faster for passing data to-from GPU or similar OpenCL devices.
• • • ClArray<T>: automatic buffer handling for C# and C++ arrays, used for computing with ClNumberCruncher
• • • ClByteArray: C++ byte array wrapper that can be used to initialize ClArray<T> or compute with Cores(explicit usage case)
• • • ClFloatArray: C++ float array wrapper that can be used to initialize ClArray<T> or copute with Cores(explicit usage case)
• • ClNumberCruncher: class for computing with arrays in ClArrays, uses Cores for an implicit usage case.
• • Cores: class for computing with arrays in ClArrays, usable directly by its compute() method, used by ClNumberCruncher for a more implicit usage.
• • Tester: for quick system tests that stresses GPUs and CPUs with opencl operations.