Ray tracing system architectures and methods

A system for 3-D graphics processing, comprising:

• a plurality of computation units, collectively capable of performing a sequence of operations using a set of data elements and outputting a respective result for each set of data elements;
• an input buffer configured for receiving data elements to be processed by the plurality of computation units;
• a controller, for the plurality of computation units, coupled to the input buffer, the controller operable to determine a sequence of sets of data elements to be inputted to the plurality of computation units, wherein
• each set of data elements is selected from among data elements received at the input buffer and data elements previously received at the input buffer and which have been added to a set and for which results have been outputted by the plurality of computation units, and
• the controller is operable to aggregate data elements received at different times into one of the sets of data elements to be provided to the plurality of computation units and to provide a maximum delay in processing each grouping of data elements received at the input buffer.

A method of graphics computation performed in a computation system, comprising:

• outputting, on a clocking schedule, a grouping of specifications of computation to be performed, a number of specifications in the grouping related to a number of specifications of computation that were available to be grouped before a clocking schedule deadline;
• receiving the outputted grouping at an input buffer to a plurality of computation units;
• re-organizing the data elements from a plurality of the groupings received at the input buffer to produce regularly sized groupings to be processed by the plurality of computation units;
• and
• processing the data elements from the regularly-sized groupings in the plurality of computation units.

A system for 3-D graphics processing, comprising:

• a plurality of computation units, collectively capable of performing a sequence of operations using a set of data elements and outputting a respective result for each set of data elements;
• an input buffer configured for receiving data elements to be processed by the plurality of computation units;
• a controller, for the plurality of computation units, coupled to the input buffer, the controller operable to determine a sequence of sets of data elements to be inputted to the plurality of computation units, wherein
• each set of data elements is selected from among data elements received at the input buffer and data elements previously received at the input buffer and which have been added to a set and for which results have been outputted by the plurality of computation units, and
• the controller is operable to aggregate data elements from groupings received at different times into one of the sets of data elements to be provided to the plurality of computation units,
• wherein the plurality of computation units are operable to process a maximum number of data elements in parallel, and the input buffer is capable of grouping data elements having a maximum number greater than the maximum number of data elements that can be processed in parallel by the plurality of computation units, and the controller is operable to delay commencement of processing of some data elements received and combine those data elements with later-received data elements to form one of the sets of data elements.