Node topology employing command and data buffers for executing thread scheduling
The described embodiments include systems, methods, and apparatuses for increased efficiency processing flow. One method includes a plurality of stages configured to process an execution graph that includes a plurality of logical nodes with defined properties and resources associated with each logical node of the plurality of logical nodes, a recirculating ring buffer, wherein the recirculating ring buffer is configured to holding only any one of a control information, input, and, or out data necessary to stream a temporary data between each logical node of the execution graph, and a data producer, wherein the data producer is configured to stall from writing control information into a command buffer upon the command buffer being full, preventing command buffer over-writing.
1. A machine configured for increased efficiency processing flow, comprising an arrangement of:
a plurality of stages configured to process an execution graph that includes a plurality of logical nodes with defined properties and resources associated with each logical node of the plurality of logical nodes;
each stage comprising:
a command buffer, wherein the command buffer comprises a recirculating ring buffer holding control information;
data buffers, wherein the data buffers comprise recirculating ring buffers holding input, and, or output data necessary to stream a temporary data between each logical node of the execution graph; and
a data producer, wherein the data producer is configured to stall from writing control information into the command buffer upon the command buffer being full, preventing command buffer over-writing, which prevents the data buffers from being over-written.
2. The machine of claim 1 , wherein the control information further comprises any of, or combination of, an execution command and, or a command related to synchronization between the logical nodes.
3. The machine of claim 1 , wherein invalidation of data of the command buffer or the data buffers is based on completion of processing a graph task.
4. The machine of claim 1 , wherein the command buffer and the data buffers that hold input data and, or out data reside in on-chip cache.
5. The machine of claim 4 , wherein the input data and, or out data is generated by the data producer and read by a consumer, whereby the data producer and consumer are proximate a memory hierarchy due to the input data and, or out data held in the data buffers residing in the on-chip cache.
6. The machine of claim 1 , wherein the commands are written into the command buffer by the data producer and the commands are read by a consumer.
7. The machine of claim 6 , wherein writing of informational control data comprises updating a write pointer in a context buffer by the data producer; and
wherein reading of informational control data comprises processing a read pointer in the context buffer by the consumer.
8. The machine of claim 7 , wherein the context buffer is pre-fetched.
9. The machine of claim 1 , wherein the command buffer and data buffers utilize addressing that wraps to a beginning of the command buffer and the data buffers once an end has been reached.
10. The machine of claim 1 , wherein a size of the command buffer and the data buffers may vary.
11. A system encoded as a set of machine-readable instructions stored on non-transitory machine-readable media, the system to employ re-circulating ring buffers to reduce computing power, said system further comprising instructions for implementing:
at least a set of logical nodes, wherein said logical nodes are processed by a plurality of stages, connected together in an arbitrary topology;
a command and data buffers residing in an on-chip cache, said command and data buffers further comprising any one of at least a control information, input and, or out data delivering temporary data between a stage and, or a logical node in an execution graph, such that any one of the control information, input, and, or out data between logical nodes need not be stored in memory;
an on-chip buffer management module, said on-chip buffer management module comprising a write pointer queue and a read pointer queue coupled via a cache gauge, thereby managing buffer content and traffic; and
based on the write pointer queue updates, direct compute resources to prepare for a data producer stall and, or buffer clear.
12. The system of claim 11 , wherein the read pointer queue alerts a cache gauge of an impending completion of a graph task and, or a maximum capacity of any one the recirculating ring buffer and the cache gauge updates the write pointer queue.
13. A non-transitory, machine-readable media, for re-configuring a machine for increased efficiency in processing flow comprising steps of:
reducing compute resources via a plurality of logical nodes being processed on plurality of stages, with defined properties and resources associated with each logical node;
reducing a size of a command buffer and data buffers by having the command buffer and the data buffers hold only any one of a control information, input, and, or out data necessary to stream a temporary data between logical nodes in an execution graph;
residing said command buffer and data buffers in an on-chip cache, such that any one of the control information, input, and, or out data between nodes need not be stored in memory; and
stalling a data producer of the machine from writing any more control information into the command buffer upon the command buffer being full, preventing command buffer over-writing.
14. The non-transitory machine of claim 13 , wherein the non-transitory machine is re-configured for increased efficiency in processing flow.