CN101976204A - Service-oriented heterogeneous multi-core computing platform and task scheduling method used by same - Google Patents
Service-oriented heterogeneous multi-core computing platform and task scheduling method used by same Download PDFInfo
- Publication number
- CN101976204A CN101976204A CN 201010506065 CN201010506065A CN101976204A CN 101976204 A CN101976204 A CN 101976204A CN 201010506065 CN201010506065 CN 201010506065 CN 201010506065 A CN201010506065 A CN 201010506065A CN 101976204 A CN101976204 A CN 101976204A
- Authority
- CN
- China
- Prior art keywords
- service
- service body
- hardware
- kernel
- calculation services
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Stored Programmes (AREA)
Abstract
The invention relates to a service-oriented heterogeneous multi-core computing platform which comprises a core servant, at least one group of software servants, at least one group of hardware servants, an interconnecting module and a peripheral bus module, wherein the core servant and the peripheral bus module are interconnected; the core servant and the hardware servants are interconnected through the interconnecting module, and the core servant and the software servants are interconnected through the interconnecting module; each software servant comprises an embedded processor and a software service program running on the embedded processor; and each hardware servant comprises a heterogeneous hardware acceleration module. The system and the scheduling method of the invention can provide various types of computing services, and adopt the hardware acceleration mechanism to reduce the computing time of the system.
Description
Technical field
The present invention relates to a kind of heterogeneous polynuclear computing platform, be specifically related to a kind of service-oriented heterogeneous polynuclear computing platform and the method for scheduling task of use thereof.
Background technology
In the ascendant to the research of chip multi-core processor (MPSoC) at present, main research focus mainly concentrates on hardware-software partition and scheduling, interconnected and aspects such as communication mechanism and design tool.It is the development trend of current embedded computing system that restructural calculating simultaneously combines with MPSoC, and this both can bring into play the computing power towards complicated applications of MPSoC, can further improve dirigibility again, satisfied multifarious computation requirement.
Service-oriented architecture (SOA) thought is a kind of emerging service-oriented thought, is mainly used in software service at present, aspects such as WEB service and operating system.It connects the different function units (being called service) of application program by definition excellent interface and contract between these services.Interface is to adopt neutral mode to define, and it should be independent of hardware platform, operating system and the programming language of the service of realization.This make the service in various such systems of being structured in can a kind of unification and universal mode carry out alternately.
The research that SOA thought is used on the MPSoC at present is also fewer, and still along with the development of restructural calculating and custom calculation, hardware system also will be close to the users more, and do corresponding configuration and optimization at specific user and application program.
Summary of the invention
One object of the present invention is to provide a kind of service-oriented heterogeneous polynuclear computing platform, another object of the present invention is to provide a kind of heterogeneous polynuclear platform employed method for scheduling task, system of the present invention and dispatching method can provide polytype calculation services, and adopt hardware-accelerated mechanism to reduce the computing time of system.
For solving technique scheme, realize above-mentioned technical purpose, the present invention is achieved through the following technical solutions:
A kind of service-oriented heterogeneous polynuclear computing platform, it comprises a kernel service body, at least one group of software service body, at least one group of hardware service body, one interconnect module, and one and the peripheral bus module that connects each other of described kernel service body, all undertaken interconnected between described kernel service body and hardware service body, kernel service body and the software service body by described interconnect module.
Further, described kernel service body comprises application programming interfaces and scheduler.
Further, described software service body is made up of described flush bonding processor and operation software service program and interface routine thereon, and described software service program is encapsulated as the form of function library.
Further, described hardware service body is made up of the hardware accelerator of isomery.
Further, described interconnect module comprises a peripheral bus, and is connected on-chip memory and at least one peripheral hardware on the described peripheral bus.
Further, current all calculation services bodies have been preserved in the described kernel service body, comprise the idle condition of described software service body and hardware service body, and be formation of calculation services body preservation of every type, be used for storing each calculating subtask of waiting for the request access services.The calculation services body is followed first the treatment principle of service earlier to the task in the formation.
The method for scheduling task that heterogeneous polynuclear computing platform of the present invention is used, it comprises step down:
Step 21) system initialization
At system's operation initial phase, described kernel service body is at first at the hardware platform in the system, carry out the initialization of the parameter of system, these parameters comprise the type and the number of body of software service described in the system and hardware service body, every kind of software service body is to all kinds task executions time and transport overhead, and every kind of hardware service body is carried out the time of specific acceleration task;
Step 22) pre-service of kernel service body application programs and analysis
When application program is moved on the heterogeneous polynuclear computing platform, described kernel service body at first carries out the pre-service of application program, analyze the part that need to obtain request calculation services body from application program, the described scheduler that application program provides by described kernel service body is realized the request visit to related service;
Step 23) kernel service body decoding
Described kernel service body whenever receives a new services request, at first obtain the operational code in the services request, be the type number of wanting requested service, obtain the requested operation number then, be the input data of request service, described kernel service body need can obtain the compute type of request by decoding, and hardware service body and software service body that this calculation services is provided;
Step 24) assignment decisions
Described kernel service body is at first checked the idle condition of current each software service body and hardware service body, if the hardware service body free time, then task scheduling is carried out to the hardware service body, otherwise carry out the comparison of software overhead and hardware spending, carry out assignment decisions then, come the judgement task on soft service body still is the hardware service body, to carry out, to obtain peak performance;
Step 25) Task Distribution is carried out to various computing service body
After obtaining the Task Distribution result, described kernel service body is transferred to the subtask in the corresponding calculation services body, and the state that upgrades corresponding calculation services body is busy state, judges that then the type of corresponding calculation services body is that obstructive type still is the unblock type;
Step 26) if the type of corresponding calculation services body is the obstructive type service, then the calculation services body in the process of implementation, described kernel service body is in waiting status, is waiting for that the calculation services body continues execution after data are returned again, and is the status indication of this calculation services body idle condition;
Step 27) if the type of corresponding calculation services body is the service of unblock type, then described kernel service body is in the process that the calculation services body that task is sent to correspondence is carried out, application program can continue to carry out current task, when the calculation services body calculating of correspondence is finished, data send back described kernel service body, comprise an interruptable controller in the described kernel service body and corresponding driver is provided, this interruptable controller detects the communication link between every kernel service body and the calculation services body, when the calculation services body carry out to finish the described kernel service body of data back, trigger the interruption of reference numeral, described kernel service body is after detecting interruption, at first interrupt the execution of current application program, and call the Interrupt Process function, the major function of Interrupt Process function is the operations such as assignment of carrying out result data, after assignment was finished, this calculation services body was idle condition.
Compared with prior art, the present invention has following significant advantage:
1, the present invention is used for the heterogeneous multi-core architecture with service-oriented thought first, has proposed heterogeneous multi-core system on the service-oriented sheet first; Improve the reusability and the extendability of system by means of the software processes device in the heterogeneous system on the sheet, and improved the performance of system by hardware accelerator.
2, the method for scheduling task of heterogeneous polynuclear computing platform use of the present invention is dispatched software service body and hardware service body simultaneously, has proposed two kinds of dissimilar implementations simultaneously for same requirements.Because carrying out efficient, hardware on specific application program, can obtain the speed-up ratio of higher degree and higher efficient than the software execution.Therefore integrated some hardware service bodies in the native system platform, and can dispatch and replace the hardware service body by the restructural technology.Thereby the hardware platform of realizing heterogeneous reconfigurable is realized the acceleration of particular task is carried out.
Above-mentioned explanation only is the general introduction of technical solution of the present invention, for can clearer understanding technological means of the present invention, and can be implemented according to the content of instructions, below with preferred embodiment of the present invention and conjunction with figs. describe in detail as after.The specific embodiment of the present invention is provided in detail by following examples and accompanying drawing thereof.
Description of drawings
Fig. 1 is the architectural schematic of an embodiment of service-oriented heterogeneous polynuclear platform of the present invention.
Fig. 2 is the schematic flow sheet of the task scheduling of service-oriented heterogeneous polynuclear computing platform of the present invention.
Embodiment
Below in conjunction with accompanying drawing concrete enforcement of the present invention is further described:
As shown in Figure 1, the organizational structure that has shown service-oriented heterogeneous polynuclear platform, described service-oriented heterogeneous polynuclear platform comprises a kernel service body 1, at least one group of software service body 2, at least one group of hardware service body 3, one interconnect module 4, and a peripheral bus module 5, all undertaken interconnected between described kernel service body 1 and hardware service body 3, kernel service body 1 and the software service body 2 by described interconnect module 4.Described heterogeneous polynuclear platform provides various service functions to application program 6.Concrete module is as follows:
1) kernel service body
Described kernel service body 1 is the important component part that constitutes heterogeneous polynuclear computing platform of the present invention.For the access interface of described application program 6 is provided simultaneously, and application programs 6 analyzes, and the subtask is dispatched, and kernel service body 1 is divided into two parts, i.e. application programming interfaces 101 and scheduler 102.Application program 6 depends on the application programming interfaces 101 that kernel service body 1 provides when design and running.And scheduler 102 is responsible for each subtask is divided on the different calculation services bodies according to the state of the attribute of task separately and current software service body 2 and hardware service body 3 and is moved.
2) software service body
Fig. 1 adopts the software task supplier of software service body 2 as system platform.Software service body 2 mainly is made of two parts: flush bonding processor 201, and the software service program 202 and the interface routine 203 of operation on described flush bonding processor 201.Each software service body 2 all operates in independently on the concurrent physical processor nuclear, and is isomorphism between the described software service body 2.Simultaneously in order to guarantee the integrality of system service, each software service body 2 can provide polytype service, all preserves the copy of the software service program that current total system can provide in each software service body 2.Described software service program can be on any one idle software service body 2 be carried out by calling of corresponding with service program in to this service function storehouse.
Further, described software service program 202 is encapsulated as the form of function library.
3) hardware service body
Described hardware service body 3 is the hardware accelerator of isomery, each hardware service body 3 provides single service, the quantity of same type of hardware accelerating module is configured according to the feature of the state of system operation and application program and optimizes, and improves the performance of system with maximum form ground.
4) interconnect module
Kernel service body 1, software service body 2, hardware service body 3 are system's abstract model among Fig. 1, in order to guarantee the normal operation of system flow, the interconnected mechanism that also needs to provide special is carried out interconnected between kernel service body 1 and software service body 2, kernel service body 1 and the hardware service body 3.Described interconnect module 4 is finished connection between the above-mentioned module by point-to-point bus.
5) peripheral bus module
Described peripheral bus module 5 comprises peripheral bus 501, on-chip memory 502 and a series of peripheral hardware 503, the described peripheral bus 501 main interconnecting functions of being responsible for kernel service body 1 and on-chip memory 502 and peripheral hardware 503.The described on-chip memory 502 main local storage of being responsible for the system-computed data, and peripheral hardware 503 is mainly used to carry out some systemic-function demonstrations and debugging etc.
Referring to shown in Figure 2, be the schematic flow sheet of the task scheduling of service-oriented heterogeneous polynuclear computing platform of the present invention.
The method for scheduling task of service-oriented heterogeneous polynuclear computing platform of the present invention may further comprise the steps:
Step 21) system initialization
At system's operation initial phase, kernel service body 1 is at first at the hardware platform in the system, carry out the initialization of the parameter of system, these parameters comprise the type and the number of software service body 2 and hardware service body 3 in the system, every kind of software service body 2 pairs of all kinds task executions time and transport overhead, every kind of time that hardware service body 3 is carried out specific acceleration task;
Step 22) pre-service of kernel service body application programs and analysis
When application program is moved on the heterogeneous polynuclear computing platform, kernel service body 1 at first carries out the pre-service of application program, analyze the part that need to obtain request calculation services body from application program, the scheduler 102 that application program provides by kernel service body 1 is realized the request visit to related service;
Step 23) kernel service body decoding
Kernel service body 1 whenever receives a new services request, at first obtain the operational code in the services request, be the type number of wanting requested service, obtain the requested operation number then, be the input data of request service, described kernel service body 1 need can obtain the compute type of request by decoding, and hardware service body 3 and software service body 2 that this calculation services is provided;
Step 24) assignment decisions
Kernel service body 1 is at first checked the idle condition of current each software service body 2 and hardware service body 3, if 3 free time of hardware service body, then task scheduling is carried out to hardware service body 3, otherwise carry out the comparison of software overhead and hardware spending, carry out assignment decisions then, come the judgement task on soft service body 2 still is hardware service body 3, to carry out, to obtain peak performance;
Step 25) Task Distribution is carried out to various computing service body
After obtaining the Task Distribution result, kernel service body 1 is transferred to the subtask in the corresponding calculation services body, and the state that upgrades corresponding calculation services body is busy, judges that then the type of corresponding calculation services body is that obstructive type still is the unblock type;
Step 26) if the type of corresponding calculation services body is the obstructive type service, then the calculation services body in the process of implementation, kernel service body 1 is in waiting status, is waiting for that the calculation services body continues execution after data are returned again, and is this calculation services body tag idle condition.
Step 27) if the type of corresponding calculation services body is the service of unblock type, then kernel service body 1 is in the process that the calculation services body that task is sent to correspondence is carried out, application program can continue to carry out current task, when the calculation services body calculating of correspondence is finished, data send back kernel service body 1, comprise an interruptable controller in the kernel service body 1 and corresponding driver is provided, this interruptable controller detects the communication link between every kernel service body 1 and the calculation services body, when the calculation services body is carried out end with data back kernel service body 1, trigger the interruption of reference numeral, kernel service body 1 is after detecting interruption, at first interrupt the execution of current application program, and call the Interrupt Process function, the major function of Interrupt Process function is the operations such as assignment of carrying out result data, after assignment was finished, this calculation services body was idle condition.
Claims (6)
1. service-oriented heterogeneous polynuclear computing platform, it is characterized in that: comprise a kernel service body (1), at least one group of software service body (2), at least one group of hardware service body (3), one interconnect module (4), and one and the peripheral bus module (5) that connects each other of described kernel service body (1), all undertaken interconnected between described kernel service body (1) and hardware service body (3), kernel service body (1) and the software service body (2) by described interconnect module (4).
2. service-oriented heterogeneous polynuclear computing platform according to claim 1 is characterized in that: described kernel service body (1) comprises application programming interfaces (101) and scheduler (102).
3. service-oriented heterogeneous polynuclear computing platform according to claim 1, it is characterized in that: described software service body (2) is made up of described flush bonding processor (201) and operation software service program (202) and interface routine (203) thereon, and described software service program (202) is encapsulated as the form of function library.
4. service-oriented heterogeneous polynuclear computing platform according to claim 1 is characterized in that: described hardware service body (3) is made up of the hardware accelerator of isomery.
5. service-oriented heterogeneous polynuclear computing platform according to claim 1, it is characterized in that: described peripheral bus module (5) comprises a peripheral bus (501), and is connected on-chip memory (502) and at least one peripheral hardware (503) on the described peripheral bus (501).
6. method for scheduling task that uses according to any described heterogeneous polynuclear computing platform in the claim 1 to 5 is characterized in that: may further comprise the steps,
Step 21) system initialization
At system's operation initial phase, kernel service body (1) is at first at the hardware platform in the system, carry out the initialization of the parameter of system, these parameters comprise the type and the number of software service body (2) and hardware service body (3) in the system, every kind of software service body (2) is to all kinds task executions time and transport overhead, and every kind of hardware service body (3) is carried out the time of specific acceleration task;
Step 22) pre-service of kernel service body application programs and analysis
When application program is moved on the heterogeneous polynuclear computing platform, kernel service body (1) at first carries out the pre-service of application program, analyze the part that need to obtain request calculation services body from application program, the scheduler (102) that application program provides by kernel service body (1) is realized the request visit to related service;
Step 23) kernel service body decoding
Kernel service body (1) whenever receives a new services request, at first obtain the operational code in the services request, be the type number of wanting requested service, obtain the requested operation number then, be the input data of request service, described kernel service body (1) need can obtain the compute type of request by decoding, and hardware service body (3) and software service body (2) that this calculation services is provided;
Step 24) assignment decisions
Kernel service body (1) is at first checked the idle condition of current each software service body (2) and hardware service body (3), if hardware service body (3) free time, then task scheduling is gone up to hardware service body (3) and carry out, otherwise carry out the comparison of software overhead and hardware spending, carry out assignment decisions then, come the judgement task in still upward execution of hardware service body (3) of soft service body (2), to obtain peak performance;
Step 25) Task Distribution is carried out to various computing service body
After obtaining the Task Distribution result, kernel service body (1) is transferred to the subtask in the corresponding calculation services body, and the state that upgrades corresponding calculation services body is busy, judges that then the type of corresponding calculation services body is that obstructive type still is the unblock type;
Step 26) if the type of corresponding calculation services body is the obstructive type service, then the calculation services body in the process of implementation, kernel service body (1) is in waiting status, is waiting for that the calculation services body continues execution after data are returned again, and is this calculation services body tag idle condition;
Step 27) if the type of corresponding calculation services body is the service of unblock type, then kernel service body (1) is in the process that the calculation services body that task is sent to correspondence is carried out, application program can continue to carry out current task, when the calculation services body calculating of correspondence is finished, data send back kernel service body (1), comprise an interruptable controller in the kernel service body (1) and corresponding driver is provided, this interruptable controller detects the communication link between every kernel service body (1) and the calculation services body, when the calculation services body is carried out end with data back kernel service body (1), trigger the interruption of reference numeral, kernel service body (1) is after detecting interruption, at first interrupt the execution of current application program, and call the Interrupt Process function, the major function of Interrupt Process function is the operations such as assignment of carrying out result data, after assignment was finished, this calculation services body was idle condition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010506065 CN101976204B (en) | 2010-10-14 | 2010-10-14 | Service-oriented heterogeneous multi-core computing platform and task scheduling method used by same |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010506065 CN101976204B (en) | 2010-10-14 | 2010-10-14 | Service-oriented heterogeneous multi-core computing platform and task scheduling method used by same |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101976204A true CN101976204A (en) | 2011-02-16 |
| CN101976204B CN101976204B (en) | 2013-09-04 |
Family
ID=43576091
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010506065 Expired - Fee Related CN101976204B (en) | 2010-10-14 | 2010-10-14 | Service-oriented heterogeneous multi-core computing platform and task scheduling method used by same |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101976204B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135950A (en) * | 2011-03-10 | 2011-07-27 | 中国科学技术大学苏州研究院 | On-chip heterogeneous multi-core system based on star type interconnection structure, and communication method thereof |
| CN102508711A (en) * | 2011-09-29 | 2012-06-20 | 中国科学技术大学苏州研究院 | Service-oriented multi-core computing platform on reconfigurable chip and reconfiguration method thereof |
| CN102890642A (en) * | 2011-12-23 | 2013-01-23 | 同济大学 | Performance analysis method based on heterogeneous reconfigurable computing (HRC) of matching matrix |
| CN111159095A (en) * | 2020-01-02 | 2020-05-15 | 中国航空工业集团公司西安航空计算技术研究所 | Heterogeneous integrated embedded intelligent computing implementation method |
| CN111274667A (en) * | 2019-12-13 | 2020-06-12 | 成都材智科技有限公司 | Cross-scale material calculation software integrated calculation system and method |
| CN113032166A (en) * | 2021-03-26 | 2021-06-25 | 黑芝麻智能科技(上海)有限公司 | Inter-core communication method, processor, inter-core communication system, and computer-readable storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101013415A (en) * | 2005-11-16 | 2007-08-08 | 阿尔卡特公司 | Thread aware distributed software system for a multi-processor array |
| US20080189714A1 (en) * | 2004-04-06 | 2008-08-07 | International Business Machines Corporation | Method, system, and storage medium for managing computer processing functions |
| CN101271401A (en) * | 2008-04-23 | 2008-09-24 | 北京航空航天大学 | Server cluster unit system with single system image |
-
2010
- 2010-10-14 CN CN 201010506065 patent/CN101976204B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080189714A1 (en) * | 2004-04-06 | 2008-08-07 | International Business Machines Corporation | Method, system, and storage medium for managing computer processing functions |
| CN101013415A (en) * | 2005-11-16 | 2007-08-08 | 阿尔卡特公司 | Thread aware distributed software system for a multi-processor array |
| CN101271401A (en) * | 2008-04-23 | 2008-09-24 | 北京航空航天大学 | Server cluster unit system with single system image |
Non-Patent Citations (1)
| Title |
|---|
| 《中国科学院研究生院学报》 20120331 王超等 异构多核平台上基于任务划分和调度的性能评估方法 第257-263页 1-6 第29卷, 第2期 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135950A (en) * | 2011-03-10 | 2011-07-27 | 中国科学技术大学苏州研究院 | On-chip heterogeneous multi-core system based on star type interconnection structure, and communication method thereof |
| CN102135950B (en) * | 2011-03-10 | 2012-11-28 | 中国科学技术大学苏州研究院 | On-chip heterogeneous multi-core system based on star type interconnection structure, and communication method thereof |
| CN102508711A (en) * | 2011-09-29 | 2012-06-20 | 中国科学技术大学苏州研究院 | Service-oriented multi-core computing platform on reconfigurable chip and reconfiguration method thereof |
| CN102890642A (en) * | 2011-12-23 | 2013-01-23 | 同济大学 | Performance analysis method based on heterogeneous reconfigurable computing (HRC) of matching matrix |
| CN102890642B (en) * | 2011-12-23 | 2014-10-22 | 同济大学 | Performance analysis method based on heterogeneous reconfigurable computing (HRC) of matching matrix |
| CN111274667A (en) * | 2019-12-13 | 2020-06-12 | 成都材智科技有限公司 | Cross-scale material calculation software integrated calculation system and method |
| CN111274667B (en) * | 2019-12-13 | 2023-06-02 | 成都材智科技有限公司 | Cross-scale material computing software integrated computing system and method |
| CN111159095A (en) * | 2020-01-02 | 2020-05-15 | 中国航空工业集团公司西安航空计算技术研究所 | Heterogeneous integrated embedded intelligent computing implementation method |
| CN111159095B (en) * | 2020-01-02 | 2023-05-12 | 中国航空工业集团公司西安航空计算技术研究所 | Heterogeneous fusion embedded intelligent computing implementation method |
| CN113032166A (en) * | 2021-03-26 | 2021-06-25 | 黑芝麻智能科技(上海)有限公司 | Inter-core communication method, processor, inter-core communication system, and computer-readable storage medium |
| CN113032166B (en) * | 2021-03-26 | 2024-05-24 | 黑芝麻智能科技(上海)有限公司 | Inter-core communication method, processor, inter-core communication system, and computer-readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101976204B (en) | 2013-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101976204B (en) | Service-oriented heterogeneous multi-core computing platform and task scheduling method used by same | |
| CN102360309B (en) | Scheduling system and scheduling execution method of multi-core heterogeneous system on chip | |
| Vaishnav et al. | Resource elastic virtualization for FPGAs using OpenCL | |
| Sokolsky et al. | Schedulability analysis of AADL models | |
| CN107077364B (en) | Compilation of graph-based program specifications using automatic clustering of graph components based on identification of specific data port connections | |
| CN100511151C (en) | Multiple-path multiple-core server and CPU virtualization processing method thereof | |
| WO2016155335A1 (en) | Task scheduling method and device on heterogeneous multi-core reconfigurable computing platform | |
| CN106598731B (en) | Runtime system and its control method based on heterogeneous polynuclear framework | |
| CN104899089A (en) | Task scheduling method in heterogeneous multi-core architecture | |
| WO2005096143A1 (en) | Resource management in a multicore architecture | |
| CN105045658A (en) | Method for realizing dynamic dispatching distribution of task by multi-core embedded DSP (Data Structure Processor) | |
| Agarwal et al. | Deadlock-free scheduling of X10 computations with bounded resources | |
| US7565659B2 (en) | Light weight context switching | |
| CN111694675A (en) | Task scheduling method and device and storage medium | |
| CN110990154A (en) | Big data application optimization method and device and storage medium | |
| CN103309734A (en) | Embedded task scheduling method based on priority grouping | |
| CN103152433A (en) | Distributed server side framework and use, interaction and request processing method thereof | |
| Brinkschulte et al. | Dynamic real-time reconfiguration in distributed systems: timing issues and solutions | |
| Bal et al. | Performance of a high-level parallel language on a high-speed network | |
| Casini et al. | Addressing analysis and partitioning issues for the Waters 2019 challenge | |
| CN103299277A (en) | Gpu system and processing method thereof | |
| CN101976206B (en) | A kind of interruption processing method and device | |
| CN101349975B (en) | Method for implementing interrupt bottom semi-section mechanism in embedded operation system | |
| Ventroux et al. | Scmp architecture: an asymmetric multiprocessor system-on-chip for dynamic applications | |
| Dziurzanski et al. | Multi-criteria resource allocation in modal hard real-time systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130904 Termination date: 20181014 |