Scheduling of scenario models for execution within different computer threads and scheduling of memory regions for use with the scenario models

A method comprising:

• receiving multiple application scenario models, wherein the application scenario models include a plurality of driver scenario models and sequences of application of the driver scenario models, wherein the application scenario models are used for testing one or more systems on chips (SoCs), wherein each application scenario model is generated by backpropagating an expected output of said each application scenario model to generate an input to be provided to said each application scenario model during a test of the one or more SoCs;
• and
• splitting multiple memory addresses into a plurality of memory regions, each of the memory regions exclusive of any other of the memory regions,
• wherein one of the driver scenario models of one of the application scenario models includes reading data from one of the memory regions, and
• wherein another one of the driver scenario models of the one of the application scenario models includes writing data to another one of the memory regions.

A system comprising:

• a memory device for storing a plurality of application scenario models, wherein the application scenario models are generated for testing one or more systems on chips (SoCs), wherein the application scenario models include a plurality of driver scenario models, wherein the application scenario models include a plurality of sequences of execution of the driver scenario models, wherein each application scenario model is generated by backpropagating an expected output of said each application scenario model to generate an input to be provided to said each application scenario model during a test of the one or more SoCs;
• and
• a processor coupled to the memory device, the processor for:
• receiving the application scenario models;
• splitting the application scenario models into a pre-determined number of computer threads, each computer thread including at least one but not all of the driver scenario models;
• splitting the sequences of execution into a plurality of flows between the driver scenario models of the computer threads;
• and
• providing, for testing of the one or more SOCs, the driver scenario models within the computer threads and the plurality of flows between the driver scenario models of the computer threads.