Texas Instruments Corporation

Sejoong Lee

Soon-Hyeok Choi

Arvind Kandhalu Raghu

Xiaolin Lu

Sandeep Bhadra

Xiaolion Lu

Karthik Singaram Lakshmanan

A multi processor task allocation method is described that considers task dependencies while performing task allocation in order to avoid blocking of a task's execution while waiting for the resolution of the dependency. While allocating the tasks to the processors the potential blocking time is considered, and the best allocation that will have the least amount of blocking time is found.

Resource sharing aware task partitioning for multiprocessors

Dynamically reducing power consumption by a processor in a computer system by determining a maximum number of times (token count) that the processor can incur a start-up delay after being placed into a low-power mode during a token period of time when executing a task for a token period of time. The processor may be placed into the low-power mode while executing the task in response to an idle indicator only if a current value of the token count assigned to the task is greater than zero. The current value of the token count is decremented each time the processor incurs a start-up delay in response to being awakened from the low-power mode. The current token count is reset to match the assigned token count at the end of each token period. Furthermore, wakeup may be anticipated to allow the processor to be awakened preemptively.

Limiting the number of unexpected wakeups in a computer system implementing a power-saving preemptive wakeup method from historical data

Systems and Methods for task allocation in a multiprocessor environment employing power management techniques are described wherein tasks are allocated relative to the density given by the ratio of worst-case-execution time and deadline of a task and also the harmonicity of a task's period with respect to a task-set. Tasks are allocated to a given processor based on either minimum density or maximum harmonicity depending on which allocation results in a lower clock frequency. Assigning a task to the processor with lowest density results in balancing the density across processors while assigning task to the processor with maximum harmonicity attempts to maximize the utilization of the processor.

System and method of task allocation in multiprocessing environment based on minimum density or maximum harmonicity that results in lower clock frequency

A method comprises operating a processor at a first power control strategy. Such a method further comprises determining whether a workload of a task running on a processor has fallen below a lower threshold, and changing to a second power control strategy to operate the processor based on determining that the workload of the task has fallen below the lower threshold.

Growing community of inventors

Allen, TX, United States of America

Sejoong Lee