CLIAMS:We claim:
1. A method for concurrently executing ordered and unordered tasks using a plurality of processing units, the method comprising:
receiving an unordered task or an ordered task into a processing queue;
processing a task in the processing queue using at least one processing unit,
wherein if the task is an unordered task or an ordered task with no other processing unit processing a related ordered task, the processing unit processes the task,
wherein if the task is an ordered task and another processing unit is processing a related ordered task, the processing unit preprocesses the ordered task if the ordered task needs preprocessing; and
dynamically managing the number of processing units.

2. The method according to claim 1, wherein ordered and unordered tasks are stored in the same processing queue.

3. The method according to claim 1, wherein the processing queue comprises a linear data structure.

4. The method according to claim 1, wherein the processing queue comprises a linear data structure, and ordered and unordered tasks are stored in the same processing queue.

5. The method according to claim 1, wherein the processing unit comprises a hardware processor.

6. The method according to claim 1, wherein the processing unit comprises a software processing unit.

7. The method according to claim 6, wherein the software processing unit comprises a thread.

8. The method according to claim 1, wherein dynamically managing the number of processing units comprises:
determining a required number of processing units based on at least one of a rate of unordered tasks being received in the processing queue, a processing rate of unordered tasks, a rate of ordered tasks being received in the processing queue, a processing rate of ordered tasks, and the number of sets of related ordered tasks in the processing queue; and
adjusting the number of processing units based on the required number of processing units.

9. The method according to claim 8, wherein the required number of processing units ( ) is equal to the lesser of or , wherein
is a maximum number of processing units,
, wherein is the rate of unordered tasks being received in the processing queue, is the processing rate of unordered tasks, and is a threshold number of processing units per unordered task, and
is determined based on values of , , and ,
wherein , wherein is a number of tasks to be preprocessed concurrently and is a threshold number of processing units per ordered task,
wherein , wherein is the rate of ordered tasks being received in the processing queue, is the processing rate of ordered tasks, and
corresponds to the number of sets of related ordered tasks in the processing queue,
and is equal to the greater of and if , or is equal to lesser of and if .

10. A system for executing ordered and unordered tasks, the system comprising one or more hardware processors and a computer-readable medium storing instructions, wherein the one or more hardware processors are configured by the instruction to:
receive an unordered task or an ordered task into a processing queue;
process a task in the processing queue using at least one processing unit,
wherein if the task is an unordered task or an ordered task with no other processing unit processing a related ordered task, the processing unit processes the task,
wherein if the task is an ordered task and another processing unit is processing a related ordered task, the processing unit preprocesses the ordered task if the ordered task needs preprocessing; and
dynamically manage the number of processing units.

11. The system according to claim 10, wherein ordered and unordered tasks are stored in the same processing queue.

12. The system according to claim 10, wherein the processing queue comprises a linear data structure.

13. The system according to claim 10, wherein the processing queue comprises a linear data structure, and ordered and unordered tasks are stored in the same processing queue.

14. The system according to claim 10, wherein the processing unit comprises a hardware processor.

15. The system according to claim 10, wherein the processing unit comprises a software processing unit.

16. The system according to claim 15, wherein the software processing unit comprises a thread.

17. The system according to claim 10, wherein dynamically managing the number of processing units comprises
determining a required number of processing units based on at least a rate of unordered tasks being received in the processing queue, a processing rate of unordered tasks, a rate of ordered tasks being received in the processing queue, a processing rate of ordered tasks, and the number of sets of related ordered tasks in the processing queue; and
adjusting the number of processing units based on the required number of processing units.

18. The system according to claim 17, wherein the required number of processing units ( ) is equal to the lesser of or , wherein:
is a maximum number of processing units,
, wherein is the rate of unordered tasks being received in the processing queue, is the processing rate of unordered tasks, and is a threshold number of processing units per unordered task, and
is determined based on values of , , and ,
wherein , wherein is a number of tasks to be preprocessed concurrently and is a threshold number of processing units per ordered task,
wherein , wherein is the rate of ordered tasks being received in the processing queue, is the processing rate of ordered tasks, and
corresponds to the number of sets of related ordered tasks in the processing queue,
and is equal to the greater of and if , or is equal to lesser of and if .

19. A non-transitory computer-readable medium storing instructions for executing ordered and unordered tasks, wherein upon execution of the instructions by one or more hardware processors, the one or more hardware processors are configured by the instructions to:
receive an unordered task or an ordered task into a processing queue;
process a task in the processing queue using at least one processing unit,
wherein if the task is an unordered task or an ordered task with no other processing unit processing a related ordered task, the processing unit processes the task,
wherein if the task is an ordered task and another processing unit is processing a related ordered task, the processing unit preprocesses the ordered task if the ordered task needs preprocessing; and
dynamically manage the number of processing units.

20. The non-transitory computer-readable medium according to claim 19, wherein the processing queue comprises a linear data structure.

21. The non-transitory computer-readable medium according to claim 19, wherein the processing queue comprises a linear data structure, and ordered and unordered tasks are stored in the same processing queue.

22. The non-transitory computer-readable medium according to claim 19, wherein dynamically managing the number of processing units comprises:
determining a required number of processing units based on at least a rate of unordered tasks being received in the processing queue, a processing rate of unordered tasks, a rate of ordered tasks being received in the processing queue, a processing rate of ordered tasks, and the number of sets of related ordered tasks in the processing queue; and
adjusting the number of processing units based on the required number of processing units.

23. The non-transitory computer-readable medium according to claim 22, wherein the required number of processing units ( ) is equal to the lesser of or , wherein:
is a maximum number of processing units;
, wherein is the rate of unordered tasks being received in the processing queue, is the processing rate of unordered tasks, and is a threshold number of processing units per unordered task; and
is determined based on values of , , and ,
wherein , wherein is a number of tasks to be preprocessed concurrently and is a threshold number of processing units per ordered task,
wherein , wherein is the rate of ordered tasks being received in the processing queue, is the processing rate of ordered tasks, and
corresponds to the number of sets of related ordered tasks in the processing queue,
and is equal to the greater of and if , or is equal to lesser of and if .

24. A method for dynamically managing processing units concurrently processing ordered tasks and unordered tasks stored in processing queue, the method comprising:
determining a required number of processing units based on at least a rate of unordered tasks being received in the processing queue, a processing rate of unordered tasks, a rate of ordered tasks being received in the processing queue, a processing rate of ordered tasks, and the number of sets of related ordered tasks in the processing queue; and
adjusting the number of processing units based on the required number of processing units.

25. The method according to claim 24, wherein the required number of processing units ( ) is equal to the lesser of or , wherein:
is a maximum number of processing units;
, wherein is the rate of unordered tasks being received in the processing queue, is the processing rate of unordered tasks, and is a threshold number of processing units per unordered task; and
is determined based on values of , , and ,
wherein , wherein is a number of tasks to be preprocessed concurrently and is a threshold number of processing units per ordered task,
wherein , wherein is the rate of ordered tasks being received in the processing queue, is the processing rate of ordered tasks, and
corresponds to the number of sets of related ordered tasks in the processing queue,
and is equal to the greater of and if , or is equal to lesser of and if .

Dated this 09th day of October, 2013

MADHUSUDAN S.T.
OF K & S PARTNERS
ATTORNEY FOR THE APPLICANTS

,TagSPECI:FIELD OF THE INVENTION
BACKGROUND
Over the past sixty years, the use of computing devices to perform large-scale calculations has become an integral force in almost all technology sectors. Over much of that time, the paradigm of computing has largely adhered to a serial execution model in which calculations are performed sequentially until a desired solution has been achieved. Archetypal problems falling into this paradigm include iterative solutions to non-linear differential equations such the determination of solutions for many-body problems in Newtonian and non-Newtonian systems. Serial execution is also the manner in which humans have naturally approached most problems including, for example, the familiar algorithms for long division or for multiplication taught to young elementary-school students. A hallmark of serial execution has been the reliance of subsequent calculations on solutions determined by earlier calculations. This reliance implies the need for serial execution because, without the solution obtained in an initial phase, a later phase is unable to begin.