Method for handling of out-of-order opaque and alpha ray/primitive intersections

A method implemented by a hardware-based traversal coprocessor coupled to a streaming multiprocessor, the method comprising:

• storing, in memory, at least a portion of an acceleration data structure including a plurality of hierarchical nodes, at least one node identifying primitives of a virtual scene;
• determine, in the order the primitives identified in the at least one node are received from memory, primitives which are intersected by a ray;
• and
• report at least a portion of the primitives the ray is determined to intersect in a memory order the primitives are stored in the memory.

A hardware-based traversal coprocessor configured to:

• receive, from a streaming multiprocessor, a query identifying a ray and a tree data structure to test for primitive intersections with the ray;
• retrieve, from memory outside of the tree traversal device, portions of the tree data structure, each portion identifying one or more bounding volumes of a hierarchy of bounding volumes identified in the tree data structure;
• determine smallest subdivisions of the hierarchy of bounding volumes which are intersected by the ray;
• push onto a stack a range of primitives identified in the smallest subdivision of the hierarchy of bounding volumes intersected by the ray;
• determine, in the order the primitives in the range of primitives are received from memory, primitives which are intersected by the ray;
• store one or more intersections for primitives determined to be intersected by the ray in a result queue in a memory order the primitives determined to be intersected by the ray are stored in the memory;
• return, to the streaming multiprocessor, intersection information of the intersected primitives stored in the result queue.

A system including:

• memory configured to store at least a portion of an acceleration data structure including a plurality of hierarchical nodes, at least one node identifying primitives of a virtual scene;
• and
• hardware-based traversal coprocessor operatively coupled to the memory and a multiprocessor, the traversal coprocessor is configured to:
• traverse the acceleration data structure to determine primitives in the at least one node which are intersected by a ray;
• and
• mid-traversal of the acceleration data structure, report, to the multiprocessor, intersection information for a primitive in the at least one node determined to be intersected by the ray.

A system including:

• memory configured to store at least a portion of an acceleration data structure including a plurality of hierarchical nodes, at least one node identifying a range of primitives of a virtual scene, the primitives of the primitive range are stored in a plurality of cache line sized blocks each including a different set of primitives of the primitive range;
• and
• autonomous hardware circuitry operatively coupled to the memory configured to:
• process the blocks, in the order the blocks are received from memory, to determine primitives intersected by a ray, wherein the blocks are received by the hardware circuitry out of order in which they are stored in the memory;
• and
• report, after each primitive of the primitive range has been processed, intersection information for at least a portion of the primitives the ray is determined to intersect.