Dynamic allocation of CPU cycles vis-a-vis virtual machines in video stream processing

A non-transitory computer-readable storage medium storing one or more sequences of instructions for dynamically allocating CPU cycles for use in processing a video stream, which when executed by one or more processors, causes:

• upon the occurrence of each periodic interval in a series of periodic intervals, performing:
• (a) determining video complexity information for two or more digital video streams actively being processed by a single video encoder;
• and (b) determining whether an amount of CPU cycles processing a particular digital video stream of the two or more digital video streams should be adjusted based on the determining video complexity information and assigned capabilities of a virtual machine in which the particular digital video stream is being processed;
• and
• dynamically increasing the amount of CPU cycles allocated to the virtual machine based on, at least in part, the determined video complexity information, wherein dynamically increasing comprising:
• assessing a first impact of decreasing the amount of CPU cycles allocated to the virtual machine processing at least one digital video stream of the two or more digital video streams, and
• assessing a second impact of increasing the amount of CPU cycles allocated to the virtual machine processing the particular digital video stream proportional to a total amount of CPU cycles reduced from said at least one digital video stream.

An apparatus for dynamically allocating CPU cycles for use in processing a video stream, comprising:

• one or more processors;
• and
• one or more non-transitory, machine-readable storage mediums storing one or more sequences of instructions, which when executed, cause:
• upon the occurrence of each periodic interval in a series of periodic intervals, performing:
• (a) determining video complexity information, for two or more digital video streams that are each actively being processed by a separate encoding module of a single video encoder, associated with one or more future intervals in said series of periodic intervals;
• and (b) determining whether an amount of CPU cycles allocated to processing a particular digital video stream of the two or more digital video streams should be adjusted in a particular future interval in said series of periodic intervals based on (i) a dynamically assessed impact on video quality in said two or more digital video streams after being encoded by said single video encoder after adjusting the amount of CPU cycles allocated to one or more separate encoding modules of said single video encoder, (ii) the determined video complexity information for the two or more digital video streams during said particular future interval, and (iii) assigned capabilities of a virtual machine in which the particular digital video stream is being processed,
• wherein each encoding module of said single video encoder is implemented solely in software and not in hardware, wherein said single video encoder comprises a number of physical CPUs, and wherein each encoding module is allocated to use a certain allotment of CPU cycles, of the number of physical CPUs, in each periodic interval of said series of periodic intervals.

An apparatus for dynamically allocating CPU cycles for use in processing a video stream, comprising:

• one or more processors;
• and
• one or more non-transitory, machine-readable storage mediums storing one or more sequences of instructions, which when executed, cause:
• upon the occurrence of each periodic interval in a series of periodic intervals, performing:
• (a) determining video complexity information for two or more digital video streams actively being processed by a single video encoder;
• and (b) determining whether an amount of CPU cycles processing a particular digital video stream of the two or more digital video streams should be adjusted based on the determining video complexity information and assigned capabilities of a virtual machine in which the particular digital video stream is being processed;
• and
• dynamically increasing the amount of CPU cycles allocated to the virtual machine based on, at least in part, the determined video complexity information, wherein dynamically increasing comprising:
• assessing a first impact of decreasing the amount of CPU cycles allocated to the virtual machine processing at least one digital video stream of the two or more digital video streams, and;
• assessing a second impact of increasing the amount of CPU cycles allocated to the virtual machine processing the particular digital video stream proportional to a total amount of CPU cycles reduced from said at least one digital video stream.

A non-transitory computer-readable storage medium storing one or more sequences of instructions for dynamically allocating CPU cycles for use in processing a video stream, which when executed by one or more processors, causes:

• executing on a physical machine a number of calibrated computer processes, each of the calibrated computer processes having a calibrated computing load defined by one or more of a CPU load, a memory load, and a bandwidth load;
• checking if an output of a computing machine matches or is better than a performance threshold defined by one or more of a quality parameter of the output and a latency parameter between an input into the computing machine and an output from the computing machine, wherein the computing machine is the physical machine or a virtual machine;
• iteratively adding or removing a calibrated computer process until a stop criterion defined by a required level of precision is achieved;
• calculating the computing capacity of the computing machine as the sum of the calibrated computer loads of the calibrated computer processes running on the computing machine when the stop criterion is achieved;
• and
• after processing a digital video stream in said computing machine, adjusting an amount of CPU cycles processing the digital video stream based on video complexity information determined for the digital video stream and the calculated computing capacity of the computing machine.