executing a shader program that performs ray tracing;

using a hardware-implemented ray tracing unit (RTU) that traverses an acceleration structure at the request of the shader program; and

using, at the shader program, results of the acceleration structure traversal;

wherein the shader program is configured to send root node and information related to rays to the RTU which is configured to traverse the acceleration structure, with the shader program being configured for reading at least one status of the RTU and the RTU reporting at least a first status as it continues its traversal, the shader program passing hit identifications to the RTU to enable the RTU to shorten rays, at least a second status from the RTU indicating that the RTU has found an intersection with a first primitive, the shader program performing hit testing and responsive to finding that the first primitive was hit by a ray, informs the RTU of the hit.