Apparatus and method for efficient scheduling of tasks
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06F-001/00
G06F-009/48
G06F-001/32
출원번호
US-0792546
(2013-03-11)
등록번호
US-9135063
(2015-09-15)
발명자
/ 주소
Ghose, Kanad
출원인 / 주소
The Research Foundation for the State University of New York
대리인 / 주소
Hoffberg, Esq., Steven M.
인용정보
피인용 횟수 :
5인용 특허 :
62
초록▼
A system and method of scheduling tasks, comprising receiving activity and performance data from registers or storage locations maintained by hardware and an operating system; storing calibration coefficients associated with the activity and performance data; computing an energy dissipation rate bas
A system and method of scheduling tasks, comprising receiving activity and performance data from registers or storage locations maintained by hardware and an operating system; storing calibration coefficients associated with the activity and performance data; computing an energy dissipation rate based on at least the activity and performance data; and scheduling tasks under the operating system based on the computed energy dissipation rate.
대표청구항▼
1. A method of scheduling tasks, comprising: receiving data conveying information relating to a plurality of prospective tasks in a processing queue;computing, with at least one automated processor, estimates of prospective energy dissipation over time for executing a load comprising the plurality o
1. A method of scheduling tasks, comprising: receiving data conveying information relating to a plurality of prospective tasks in a processing queue;computing, with at least one automated processor, estimates of prospective energy dissipation over time for executing a load comprising the plurality of tasks, based on at least the received data and predetermined calibration parameters for associating respective tasks with a related energy dissipation over time; andselectively scheduling the plurality of prospective tasks for processing by at least one resource which each has a respective temperature determined by different factors, based on at least the computed estimate of energy dissipation affected by the scheduled tasks, and at least a predicted load-dependent temperature associated with each at least one resource over time. 2. The method of claim 1, wherein the scheduling of the plurality of prospective tasks is further dependent on a predicted system load and the plurality of prospective tasks are distributed unevenly to the at least one resource. 3. The method of claim 1, wherein the scheduling of the plurality of prospective tasks is dependent on a predicted load-dependent temperature associated with each of at least two resources over time. 4. The method of claim 1, wherein the scheduling of the plurality of prospective tasks is based on at least the computed estimate of energy dissipation affected by the scheduled tasks, a predicted system load, and a predicted load-dependent temperature associated with each at least two resources over time. 5. The method of claim 1, further comprising coordinating a control of a dynamically controllable cooling system with the scheduling, the dynamically controllable cooling system being effective for varying a relationship between an aggregate energy dissipation over time and temperature of the at least one resource. 6. The method of claim 1, wherein the scheduling comprises reordering a plurality of elements in the processing queue. 7. The method of claim 1, further comprising selectively scheduling the plurality of prospective tasks for processing by at least one resource to meet at least one service guaranty. 8. The method of claim 1, further comprising recruiting at least one inactive resource to become available for executing tasks in dependence on a predicted temperature associated with at least one other resource, such that the predicted load dependent temperature associated with the at least one other resource does not exceed a maximum operating temperature. 9. An apparatus for scheduling tasks, comprising: a processing queue configured to receive data conveying information relating to a plurality of prospective tasks;at least one automated processor, configured to compute estimates of prospective energy dissipation over time for executing a load comprising the plurality of tasks based on at least the received data and predetermined calibration parameters for associating respective tasks with a related energy dissipation over time; anda scheduler configured to schedule the plurality of prospective tasks for processing by at least one resource having a respective temperature determined by different factors, based on the computed estimate of energy dissipation affected by the scheduled tasks, and a predicted load-dependent temperature associated with each of the at least one resource over time. 10. The apparatus of claim 9, wherein the scheduler is configured to schedule the plurality of prospective tasks in dependence on at least a predicted system load, such that the plurality of prospective tasks are distributed unevenly to the at least one resource. 11. The apparatus of claim 9, wherein the scheduler is configured to schedule the plurality of prospective tasks in dependence on a predicted load-dependent temperature associated with each of at least two resources over time. 12. The apparatus of claim 9, wherein the scheduler is configured to schedule the plurality of prospective tasks in dependence on at least the computed estimate of energy dissipation affected by the scheduled tasks, a predicted system load, and a predicted load-dependent temperature associated with each of at least two resources over time. 13. The apparatus of claim 9, further comprising a control for dynamically controlling a cooling system, configured to coordinate operation of the dynamically controllable cooling system with the schedule of the plurality of tasks, and thereby selectively influence both an allocation of the plurality of prospective tasks to a respective at least one resource and a cooling of the respective at least one resource. 14. The apparatus of claim 9, wherein each respective task has a classification based on of a plurality of different classifications based on usage characteristics of the at least one resource, and the scheduler is configured to selectively schedule respective tasks having the same classification back to back. 15. The apparatus of claim 9, wherein the scheduler is further configured to selectively schedule the plurality of prospective tasks for processing by at least one resource to meet at least one service guaranty. 16. The apparatus of claim 9, further comprising at least one inactive resource, the scheduler being further configured to activate the at least one inactive resource to become available for executing tasks in dependence on a predicted temperature associated with at least one active resource, such that the predicted load dependent temperature associated with the at least one active resource does not exceed a maximum operating temperature. 17. A method of scheduling tasks, comprising: maintaining a queue of tasks to be prospectively processed by a system having at least two processing resources;estimating an energy dissipation over time for processing the respective tasks in the queue based on previous energy dissipation for corresponding tasks; andreordering tasks in the queue of tasks for selective dispatch to each of the at least two processing resources, based on at least the estimated energy dissipation over time affected by the scheduled tasks, and a predicted load-dependent temperature associated with each resource over time. 18. The method of claim 17, wherein the allocating is further dependent on a predicted system load and at least one quality of service constraint. 19. The method of claim 17, further comprising coordinating a control of a dynamically controllable cooling system with the allocating, the dynamically controllable cooling system being effective for varying a relationship between an aggregate energy dissipation over time and temperature of the at least one resource. 20. The method of claim 17, further comprising activating at least one additional resource to process tasks in dependence on at least the estimated energy dissipation over time affected by the scheduled tasks, such that the predicted load dependent temperature associated with the at least two active resources does not exceed a respective maximum operating temperature. 21. The method of claim 17, wherein each respective task has a classification based on of a plurality of different classifications based on usage characteristics of the at least one resource, and the reordering of tasks in the queue of tasks for dispatch comprises reordering the tasks for dispatch to each least two alternate resources the scheduler to selectively schedule respective tasks having the same classification back to back on the same respective at least one resource.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (62)
Hass, David T., Advanced processor with mechanism for packet distribution at high line rate.
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Apparatus for compact internetworked wireless integrated network sensors (WINS).
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Apparatus for internetworked hybrid wireless integrated network sensors (WINS).
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Apparatus for internetworked wireless integrated network sensors (WINS).
Tang Jun ; So John Ling Wing, Computer operating process allocating tasks between first and second processors at run time based upon current processor load.
So John Ling Wing ; Kerr Jeffrey L. ; Magee Steven R. ; Tang Jun, Devices, methods, systems and software products for coordination of computer main microprocessor and second microprocessor coupled thereto.
Titiano, Patrick C; Qasem, Safwan, Hardware voting mechanism for arbitrating scaling of shared voltage domain, integrated circuits, processes and systems.
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Method and apparatus for distributed signal processing among internetworked wireless integrated network sensors (WINS).
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Method and apparatus for internetworked wireless integrated network sensor (WINS) nodes.
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Method for collecting data using compact internetworked wireless integrated network sensors (WINS).
Gelvin, David C.; Girod, Lewis D.; Kaiser, William J.; Merrill, William M.; Newberg, Fredric; Pottie, Gregory J.; Sipos, Anton I.; Vardhan, Sandeep, Method for internetworked hybrid wireless integrated network sensors (WINS).
Yamazaki, Takeshi; Brown, Jeffrey Douglas; Clark, Scott Douglas; Johns, Charles Ray, Methods and apparatus for facilitating coherency management in distributed multi-processor system.
Barth, Andy; Dall'Acqua, Daniele; Drew, Nigel, Performance estimation for adjusting processor parameter to execute a task taking account of resource available task inactive period.
Catthoor Francky,BEX ; Miranda Miguel,BEX ; Janssens Stefan,BEX ; De Man Hugo,BEX, Power-and speed-efficient data storage/transfer architecture models and design methodologies for programmable or reusable multi-media processors.
Ghose,Kanad; Ponomarev,Dmitry V.; Kucuk,Gurhan, System and method for reducing power requirements of microprocessors through dynamic allocation of datapath resources.
Li, Cheng-Wei; Shen, Chung-Chou, System having minimum latency using timed mailbox to issue signal in advance to notify processor of the availability of the shared resources.
William N. Joy ; Marc Tremblay ; Gary Lauterbach ; Joseph I. Chamdani, Vertically and horizontally threaded processor with multidimensional storage for storing thread data.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.