Method and apparatus for the allocation of audio/video tasks in a network system

A method for performing audio/video processing comprising:

• inputting at least one input task to be performed;
• determining usage requirements of the at least one input task by referencing a process descriptor block configured to identify usage requirements of tasks, said usage requirements comprising those requirements to be supported by support resources, including a plurality of digital signal processors (DSPs) and at least one device, at least some of the DSPs and at least one device coupled together via a network;
• determining available bandwidth of digital signal processors of the plurality of DSPs;
• determining expandability of the at least one input task to networked devices;
• determining the availability of the at least one device;
• allocating the input task to support resources based upon the determined usage requirements, the determined available bandwidth, the expandability of the task across the network to networked devices, and the determined availability of the at least one device;
• and
• selectively configuring each DSP of the allocated support resources to perform the input task.

A method for distributing an audio/video task comprising:

• receiving the audio/video (A/V) task;
• segmenting the A/V task into at least one primitive, each of the at least one primitive representing a basic processing component of the task;
• determining at least one support resource required to execute each of the at least one primitive by accessing requirement information relating to the capacity necessary to execute a set of predefined primitive functions, the requirement information including bandwidth needed for each of the at least one primitive, support resources needed for the at least one primitive, and expandability of the at least one primitive across at least one of the at least one DSP and networked support resources;
• maintaining capacity information indicative of current processing capacity of support resources available to execute at least one A/V task comprising corresponding primitives;
• comparing the requirements information with the configuration information;
• allocating the one or more primitives to one or more support resources based on the comparison;
• and
• configuring each allocated DSP of the allocated support resources to execute a corresponding, identified primitive.

A scalable audio/video processor system comprising:

• support resources to process A/V (audio/visual) tasks, said support resources including a plurality of configurable digital signal processors (DSPs);
• a process descriptor block to identify usage requirements of at least one task primitive, said usage requirements for a given primitive including processing requirements of the given primitive that are to be supported by the support resources;
• and
• a control processor coupled to the support resources and to the process descriptor block, said control processor configured to:
• segment an input A/V task into at least one primitive;
• determine the usage requirements for each of the at least one segmented primitives by reference to the usage requirements in the process descriptor block;
• providing a time stamp to each of the one or more segmented primitives, the timestamp indicating when execution of a given segmented primitive is to be started and when the output of the given segmented primitive is to be generated by;
• determining the availability of the support resources;
• if each of the at least one segmented primitives can be executed within their corresponding start and output times, then allocating each of the at least one segmented primitives to selected support resources based upon the determined usage requirements;
• control the configuration of each DSP of the allocated support resources to perform one or more segmented primitives as allocated.