Uniform predicates in shaders for graphics processing units

摘要

A method for processing data in a graphics processing unit includes receiving an indication that all threads of a warp in a graphics processing unit (GPU) should execute the same branch in a first set of instructions, receiving a single set of predicate bits The method comprising the steps of: storing one or more predicate bits in a memory, wherein a single set of predicate bits is applied to all threads in a warp; storing the one or more predicate bits in a single set of predicate bits; And executing a portion of the first set of instructions. The execution of the first set of instructions may be performed by executing a first set of instructions in accordance with a single set of predicate bits using a single instruction, multiple data (SIMD) processing core and / or using a scalar processing unit to generate a predicate bit And executing the first set of instructions according to a single set of instructions.