Abstract: A method of generating a shader program includes determining a first variable to be used to perform a folding operation that replaces operation formulas included in the shader program by a constant or a second variable, and generating a shader program in which the operation formulas are replaced by the constant or the second variable based on the determined first variable.

Abstract: Methods and apparatuses for determining a work-group size are provided. The method of determining a work-group size includes calculating a first value denoting spatial locality of a memory that is shared by one or more work items included in a work-group, calculating a second value denoting footprints of the one or more work items included in the work-group based on the first value, and determining the work-group size based on the first and second values.

Abstract: In a pipelined application having different stages of processing, such as a graphics application or an image processing application, there may be a dependence of one compute kernel upon another. Data associated with individual kernels needs to be written and read. A technique to minimize a need to read and write kernel data to external memory utilize at least one of fusing kernels, resizing workgroups, and performing interleaving of kernels.

Abstract: A method of generating a shader program includes determining a first variable to be used to perform a folding operation that replaces operation formulas included in the shader program by a constant or a second variable, and generating a shader program in which the operation formulas are replaced by the constant or the second variable based on the determined first variable.

Abstract: In a pipelined application having different stages of processing, such as a graphics application or an image processing application, there may be a dependence of one compute kernel upon another. Data associated with individual kernels needs to be written and read. A technique to minimize a need to read and write kernel data to external memory utilize at least one of fusing kernels, resizing workgroups, and performing interleaving of kernels.

Abstract: Methods and apparatuses for determining a work-group size are provided. The method of determining a work-group size includes calculating a first value denoting spatial locality of a memory that is shared by one or more work items included in a work-group, calculating a second value denoting footprints of the one or more work items included in the work-group based on the first value, and determining the work-group size based on the first and second values.

Abstract: A method of processing data includes classifying input data into first data and second data, the second data being different from the first data, separately compiling the first data and the second data, and providing the compiled first data and the compiled second data to a first operator and a second operator, respectively, in which the first operator performs an operation different from an operation performed by the second operator.

Abstract: A memory system includes a central processing unit (CPU), a nonvolatile memory electrically coupled to the CPU and a main memory, which is configured to swap an incoming code page for a target code page therein, in response to a first command issued by the CPU. The main memory can be configured to swap the target code page in the main memory to the nonvolatile memory in the event a page capacity of the main memory is at a threshold capacity. The CPU may also be configured to perform a frequency of use analysis on the target code page to determine whether the target code page is to be swapped to the nonvolatile memory or discarded. The incoming code page may be provided by a disk drive storage device and the main memory may be a volatile memory.

Abstract: A memory system includes a central processing unit (CPU), a nonvolatile memory electrically coupled to the CPU and a main memory, which is configured to swap an incoming code page for a target code page therein, in response to a first command issued by the CPU. The main memory can be configured to swap the target code page in the main memory to the nonvolatile memory in the event a page capacity of the main memory is at a threshold capacity. The CPU may also be configured to perform a frequency of use analysis on the target code page to determine whether the target code page is to be swapped to the nonvolatile memory or discarded. The incoming code page may be provided by a disk drive storage device and the main memory may be a volatile memory.