Scheduling and execution of compute tasks

A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to assigning a compute task to a first processor included in a plurality of processors, by performing the steps of:

• analyzing each compute task in a plurality of compute tasks to identify one or more compute tasks that are eligible for assignment to the first processor,
• wherein the first processor comprises a single-instruction-multiple-data (SIMD) processor that includes multiple execution units, and a thread group having a plurality of threads executes across the multiple execution units to process different data via sets of substantially identical instructions, wherein each compute task is listed in a first table and is associated with a priority value and an allocation order that indicates a time at which the compute task was added to the first table;
• selecting a first compute task from the identified one or more compute tasks based on at least one of the priority value and the allocation order;
• assigning the first compute task to the first processor for execution;
• spawning one or more cooperative thread arrays (CTAs) that includes the thread group;
• and
• executing the one or more CTAs.

A system for assigning a compute task to a first processor included in a plurality of processors, the system comprising:

• a memory that is configured to a plurality of compute tasks;
• the plurality of processors;
• and
• a scheduling unit that is configured to:
• analyze each compute task in a plurality of compute tasks to identify one or more compute tasks that are eligible for assignment to the first processor, wherein the first processor comprises a single-instruction-multiple-data (SIMD) processor that includes multiple execution units, and a thread group having a plurality of threads executes across the multiple execution units to process different data via sets of substantially identical instructions, wherein each compute task is listed in a first table and is associated with a priority value and an allocation order that indicates a time at which the compute task was added to the first table;
• select a first compute task from the identified one or more compute tasks based on at least one of the priority value and the allocation order;
• and
• assign the first compute task to the first processor for execution,
• wherein the plurality of processors are configured to execute one or more cooperative thread arrays (CTAs) that includes the thread group.