CPU container for a tensor field, including methods for reading and writing it.
CPU container for a tensor field, including methods for reading and writing it. This implemented using direct, NIO buffers for efficiency.
This organizes the data within the buffer so that memory accesses by the GPU are efficient. This also provides a uniform representation of fields that's consistent on both CPU and GPU.
See the FieldMemoryLayout class for description of the physical layout.
CPU memory for an image.
CPU memory for an image.
Internally this assumes RGBA format, which is the only format guaranteed to be supported by OpenCL.
CPU memory for a complex scalar field.
CPU memory for a complex vector field.
A class for producing a field memory from a field type.
CPU memory for a matrix field.
CPU memory for a scalar field.
CPU memory for a vector field.
Factory for creating ColorFieldMemories.
Factory for creating ColorFieldMemories.
THIS IS ONLY USED FOR TESTING. NOT EXPORTED TO USERS.
Factory for creating ComplexFieldMemories.
Factory for creating ComplexFieldMemories.
THIS IS ONLY USED FOR TESTING. NOT EXPORTED TO USERS.
A factory for producing field memory allocators.
A factory for producing field memory allocators.
Currently each ComputeGraph gets its own allocator. Some uses within the Cogdebugger didn't have access to the ComputeGraph's allocator and so use a global one.
Factory for creating MatrixFieldMemories.
Factory for creating MatrixFieldMemories.
THIS IS ONLY USED FOR TESTING. NOT EXPORTED TO USERS.
Factory for creating ScalarFieldMemories.
Factory for creating ScalarFieldMemories.
THIS IS ONLY USED FOR TESTING. NOT EXPORTED TO USERS.
Factory for creating VectorFieldMemories.
Factory for creating VectorFieldMemories.
THIS IS ONLY USED FOR TESTING. NOT EXPORTED TO USERS.
This package contains classes for holding fields in CPU memory. The format of fields is hidden and identical to format used to hold fields in GPU memory. "Direct memory" is used to hold the field data since it exists outside of the JVM heap, and also handles all endian-ness compatibility issues.