Abstract:
A method of processing ray data the method including receiving ray data stored in input buffers; updating a property of the ray data based on a result of intersection traversal of the ray data; and outputting the ray data to an output buffer or to a feedback buffer from among the input buffers based on the updated property of the ray data. When there is insufficient space in the output buffer to store the ray data, outputting the ray data to the feedback buffer.
Abstract:
An apparatus and method for detecting an error occurring when an application program is executed in a computer environment is provided. The error detection apparatus may measure a deterministic progress index (DPI) and a program counter (PC) value when an instruction is executed, set, as a verification set, a DPI and a PC value measured when the instruction is executed without causing an error, set, as a measurement set, the DPI and the PC value measured when an instruction is executed, and detect a runtime error of the instruction by comparing the measurement set to the verification set.
Abstract:
A processing apparatus for managing power based on data is provided. The processing apparatus may obtain, in response to an access request from a processor for particular data stored in a memory, existing power information having a predefined correspondence to the particular data, and control a power mode of the processor based on the existing power information.
Abstract:
A method and apparatus for managing a stack used for a tree traversal (TRV) may include a processing unit that may manage a short stack for a tree TRV, and a storage unit that may store data of the short stack. The processing unit may perform a restart of the tree TRV from an intermediate node of a tree to fill the short stack when the short stack is empty.
Abstract:
A graphic processing apparatus and method for processing ray tracing may include a plurality of traversal units to process traversal of a ray. A management unit of the graphics processing apparatus may distribute data of the ray processed by the graphics processing apparatus to the plurality of traversal units. Each of the plurality of traversal units may process ray traversal with respect to a subdivision of the entire space.
Abstract:
A virtual architecture generating apparatus and method, a runtime system, a multi-core system, and methods of operating the runtime system and the multi-core system may include analyzing a requirement of an application, a feature of the application, and a requirement of a system enabling an execution of the application, and include generating a virtual architecture corresponding to the application, based on a physical architecture of a reconfigurable processor, the analyzed requirements and the analyzed feature.
Abstract:
Provided is an apparatus and method for managing power based on data. The apparatus may include a code segment searching unit configured to search for at least one code segment in which a power type is inserted, a block determining unit configured to determine at least one block based on the at least one found code segment, and a power mode control unit configured to control the at least one determined block to operate in a power mode corresponding to the power type.
Abstract:
An apparatus and method for processing rendering data that may group vertex data that is received from a host computer, and may assign a shader processing unit to process a vertex shader among one or more shader processing units, and process the grouped vertex data using the assigned shader processing unit.
Abstract:
An apparatus and method for traversing a hierarchical acceleration structure may determine whether a current traversal node is a leaf node, may calculate a first distance from the current traversal node to a pop level and a second distance from a root node to the pop level when the current traversal node is the leaf node, and may determine a hierarchical traversal restarting position by comparing the first distance and the second distance.
Abstract:
A memory overlay apparatus includes an internal memory that includes a dirty bit indicating a changed memory area, a memory management unit that controls an external memory to store only changed data so that only data actually being used by a task during overlay is stored and restored, and a direct memory access (DMA) management unit that confirms the dirty bit when the task is changed and that moves a data area of the task between the internal memory and the external memory.