Clock acceleration of CPU core based on scanned result of task for parallel execution controlling key word
In a computing system having a multi-core central processing unit (CPU) having at least two cores, it is determined that a task to be scheduled meets clock acceleration criteria such as requiring a number of threads less than or equal to a pre-defined threshold and/or having tasks that will run above a pre-determined amount of time. Thereafter, a clock speed of a first core of the CPU is increased and a clock speed of a second core of the CPU is decreased. Once the clock speeds have been adjusted, the task is scheduled to be executed by the first core. Related apparatus, systems, techniques and articles are also described.
1. A method for implementation on a computing system having a multi-core central processing unit (CPU) having at least two cores, the method comprising:
determining, by a software implemented task scheduler executing at an application layer, that a task to be scheduled meets clock acceleration criteria, each task comprising at least one thread;
calling, by the software implemented task scheduler, an interface of an operating system of the computing system to increase a clock speed of a first core of the CPU and to decrease a clock speed of a second core of the CPU; and
scheduling the task to be executed by the first core;
wherein:
the task is encoded with at least one key word, and
the determining that the task meets the clock acceleration criteria comprises scanning the task to identify key words that control parallel execution.
2. A method as in claim 1 , wherein the clock acceleration criteria requires that the task have a number of threads less than or equal to a pre-defined threshold.
3. A method as in claim 2 , wherein the pre-defined threshold is one thread.
4. A method as in claim 1 , further comprising:
identifying an attribute associated with the task which indicates that it will run above a pre-defined amount of time;
wherein the clock acceleration criteria requires that at least one thread of the task will run longer than the pre-defined amount of time.
5. A method as in claim 1 , wherein the second core of the CPU is stopped so that the clock speed of the second core of the CPU is zero or substantially zero.
6. A method as in claim 1 , further comprising:
initiating execution of the task by the first core.
7. A method as in claim 1 , wherein the CPU dynamically assigns the first core and the second core based on processing states of such cores.
8. An article of manufacture comprising:
computer executable instructions stored on non-transitory computer readable media, which, when executed by a computer, causes the computer to perform operations comprising:
determining, by a software implemented task scheduler executing at an application layer, that a task to be scheduled meets clock acceleration criteria, each task comprising at least one thread;
calling, by the software implemented task scheduler, an interface of an operating system of the computing system to increase a clock speed of a first core of the CPU and to decrease a clock speed of a second core of the CPU; and
scheduling the task to be executed by the first core;
wherein:
the task is encoded with at least one key word, and
the determining that the task meets the clock acceleration criteria comprises scanning the task to identify key words that control parallel execution.
9. An article as in claim 8 , wherein the clock acceleration criteria requires that the task have a number of threads less than or equal to a pre-defined threshold.
10. An article as in claim 9 , wherein the pre-defined threshold is one thread.
11. An article as in claim 8 , further comprising:
identifying an attribute associated with the task which indicates that it will run above a pre-defined amount of time;
wherein the clock acceleration criteria requires that at least one thread of the task will run longer than the pre-defined amount of time.
12. An article as in claim 8 , wherein the second core of the CPU is stopped so that the clock speed of the second core of the CPU is zero or substantially zero.
13. An article as in claim 8 , wherein the CPU dynamically assigns the first core and the second core based on processing states of such cores.
14. A method for implementation on a computing system having a multi-core central processing unit (CPU) having at least two cores, the method comprising:
scanning, by a software implemented task scheduler executing at an application layer, a code of a coded task to identify at least one key word to control parallel execution, each task comprising at least one thread;
calling, by the software implemented task scheduler and based on the at least one identified key word, an interface of an operating system of the computing system to increase a clock speed of a first core of the CPU and to decrease a clock speed of a second core of the CPU; and
scheduling the task to be executed by the first core.