CN103257904B - Optimize computing flow graph mapping method and the device of many core system repairing performances - Google Patents
Optimize computing flow graph mapping method and the device of many core system repairing performances Download PDFInfo
- Publication number
- CN103257904B CN103257904B CN201310144403.9A CN201310144403A CN103257904B CN 103257904 B CN103257904 B CN 103257904B CN 201310144403 A CN201310144403 A CN 201310144403A CN 103257904 B CN103257904 B CN 103257904B
- Authority
- CN
- China
- Prior art keywords
- reliability
- mapping
- node
- priority
- task
- 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.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Hardware Redundancy (AREA)
Abstract
The embodiment of the present invention provides a kind of computing flow graph mapping method and device of optimizing many core system repairing performances. The method mainly comprises: for many core system computing flow graphs, for the task of each node is carried out Calculation of Reliability; Ensure that according to reliability priority and many core architecture resources generate the idle check figure order of the required vicinity taking of node; Ensure that according to the reliability obtaining priority list and system computing flow graph complete the mapping of computing flow graph to the many core frameworks of target. The method that multiple compute node is ensured in many core frameworks to priority level initializing layout by reliability that the present invention proposes, can solve preferably time delay and the power problems of many core system failures selfreparing and repair after hydraulic performance decline problem.
Description
Technical field
The present invention relates to many-core processor field, particularly a kind of computing flow graph mapping method and device of optimizing many core system repairing performances.
Background technology
At present, along with being showing improvement or progress day by day of semiconductor technology, integrated level is more and more higher, and door number that can be integrated in unit are is more and more. In recent years, industry designer has recognized that the design of processor chips is faced with two problems: the one, and Andrea Pollack rule proposes, the lifting of one single chip operation efficiency is only proportional to the square root of design complexities (door number), and the design complexities of the increase one single chip of meaning lifting operational efficiency has reached bottleneck; The 2nd, the concern of system reliability is increased day by day, chip when in use, a primary element exclusive disjunction cell failure all likely causes whole system collapse, how a System on Chip/SoC is recovered from fault, and the life-span of improving system also becomes a problem.
Monokaryon system not only, on performance improves, on the complicated energy consumption problem bringing of design, or in the treatment measures of reply reliability failure, all cannot meet designer and user's requirement. Increase the quantity of core, multinuclear cooperation parallel processing can improve system task treatment effeciency; On the other hand, many core system architectures have the resource of numerous redundancies, for the self-regeneration of processor system provides condition. Fig. 1 has shown a kind of many core frameworks of rectangular configuration, has core, route and the communication link wiring of numerous redundancies.
Existing technology mainly contains two kinds: a kind of for after breaking down when the some monokaryons of detection, can under the cooperation of controlling core, restart to reach the object of recovery; Another for when detect some monokaryons break down and restart after when also cannot recovering, described certain core is comformed in core system and is peeled off, utilize the business of the check system of redundancy to redistribute, with recovery system function.
Above-mentioned existing two kinds of shortcomings that improve the method for the reliability of many-core processor are: although the soft error that the signal intensity that the first scheme can be recovered to be caused by environment, high energy particle etc. produces, if produce fault on this core hardware, is restarted and cannot be solved; Although first scheme provides the recovery scheme that a kind of reliability is high, but this scheme does not provide a kind of mapping method that can improve self-healing properties, if there is fault in some monokaryons, and around this core, there is no idle core, this core task will be reassigned to the position far away apart from original position, or multiple core tasks will move, cause repairing rear systematic function and decline excessive and the long power consumption of selfreparing time delay is excessive, repeatedly after fault selfreparing, impact is larger.
For this reason, how the computing flow graph of a system is carried out to many core framework mappings, make fault selfreparing time delay short, low on systematic function impact after reparation, becoming one has problem to be solved. The present invention is directed to the problems referred to above, under the subsidy of problem that is numbered 513080102, proposed a kind of computing flow graph mapping method and device of optimizing many core system repairing performances.
Summary of the invention
The present invention is directed to prior art above shortcomings, a kind of computing flow graph mapping method of optimizing many core system repairing performances is provided, its object is to solve many core systems because after fault, extend degradation problem under systematic function after selfreparing when selfreparing.
The present invention is achieved through the following technical solutions:
A computing flow graph mapping method of optimizing many core system repairing performances, comprising:
For the many core system computing flow graphs that distributed task, carry out Calculation of Reliability for the task of each node, the reliability that generates each node task ensures priority;
Ensure that based on reliability priority generates the idle check figure order of the required reserved vicinity of node;
Ensure that based on reliability priority list and system computing flow graph complete the mapping of computing flow graph to the many core frameworks of target.
Preferably, for the many core system computing flow graphs that distributed task, carry out Calculation of Reliability for the task of each node, also comprise:
Task is important, and reliability ensures that priority is higher;
Task probability of malfunction is higher, and reliability ensures that priority is higher;
Or, consider this two aspects factor.
Preferably, ensure that according to reliability priority generates the idle check figure order of the required vicinity taking of node method, also comprises:
Generate contiguous idle core and need consider task node sum and many core architecture resources, different priorities distributes contiguous idle check figure order automatically to be generated or user's manual definition by the method, is kept at reliability and ensures in priority list; Many core architecture resources are configured by user.
Preferably, contiguous idle core, is characterized in that:
The scope of contiguous idle core is defined by the user.
Preferably, it is characterized in that, ensure that according to reliability priority list and system computing flow graph complete the mapping of computing flow graph to the many core frameworks of target, comprise step:
The first step, chooses not reliability in mapping node and ensures the highest node of priority;
Whether bis-Walk, there is unallocated core on the many core frameworks of query aim, carries out tri-Walk, otherwise carry out Wu Walk if exist;
The 3rd Walk, whether inquiry exists a panel region to meet the contiguous idle check figure requirement of node, carries out the 4th step, otherwise carry out the 6th step if do not exist;
The 4th step, reduces contiguous idle check figure requirement, re-starts the 3rd step;
The 5th step, choose reliability in mapping node ensure priority minimum and the node that has contiguous idle check figure, using its idle core as mapping target;
The 6th step, arrives this region by node mapping;
The 7th step, it is complete whether inspection shines upon, and carries out the 8th step, otherwise get back to the first step if complete;
The 8th step selects a core to be configured as initial position according to certain method in each node region, completion system configuration;
The 9th step, all the other do not configure core as idle spare core.
The present invention also provides a kind of computing flow graph mapping device of optimizing many core system repairing performances, and its object is to solve many core systems because after fault, extend degradation problem under systematic function after selfreparing when selfreparing.
A computing flow graph mapping device of optimizing many core system repairing performances, comprising:
Reliability prediction module, carries out Calculation of Reliability for each the node task to system computing flow graph, and generates reliability guarantee priority list;
Priority memory cell, ensures priority list for memory reliability;
Mapping dispensing unit, for ensureing that according to reliability priority list is mapped on the many core frameworks of target by system computing flow graph, completion system configuration;
Mapping area memory cell, for storing the mapping area of each task node.
Preferably, reliability prediction module, according to the importance of this task and probability of malfunction size, determines the priority size of this task automatically;
Task is important, and reliability ensures that priority is higher; Task probability of malfunction is higher, and reliability ensures that priority is higher; Or, consider this two aspects factor.
Preferably, reliability prediction module, the priority list of generation is reserved contiguous idle check figure order with needs and is represented, and contiguous idle check figure order is automatically generated or is configured by user by the computing flow graph mapping device of optimizing many core system repairing performances.
Preferably, priority memory cell, in the time that mapping distribution cannot complete, amendment is to reduce the idle check figure order of required reserved vicinity.
Preferably, mapping dispensing unit, also, in the time that mapping distribution cannot complete, proposes amendment request to priority memory cell.
Preferably, mapping area memory cell, also for when the system selfreparing, instructs the priority of task that needs to repair to be configured in the region that this core distributes.
Computing flow graph mapping method and the device of the many core systems of the optimization repairing performance that the application embodiment of the present invention provides, due to for the low compute node of reliability, reserve in its vicinity a certain amount of idle core, make in the time of system failure selfreparing, can find fast and repair place at fault near zone, complete selfreparing, reduce and repair time delay; And due to less to whole system relative position variation on many core frameworks, greatly reduce the impact of selfreparing on systematic function; Especially for a fragile task node, repeatedly fault selfreparing is more effective.
Brief description of the drawings
Shown in Fig. 1 is 4 × 4 processor array frameworks;
Fig. 2 is a kind of computing flow graph mapping method flow chart of optimizing many core system repairing performances that the embodiment of the present invention one provides;
Fig. 3 is that a kind of reliability that the embodiment of the present invention one provides ensures priority list schematic diagram;
Fig. 4 is the concrete steps flow chart of computing flow graph mapping configuration step in a kind of computing flow graph mapping method flow process of optimizing many core system repairing performances that the embodiment of the present invention one provides;
Fig. 5 is the computing flow graph of a kind of system of providing of the embodiment of the present invention one;
Fig. 6 is that the node reliability that the embodiment of the present invention one provides ensures that priority proceeds to the schematic diagram of the many core framework mappings of target configuration;
Fig. 7 is the concrete structure figure of a kind of computing flow graph mapping device of optimizing many core system repairing performances that the embodiment of the present invention two provides.
Detailed description of the invention
Below with reference to accompanying drawing of the present invention; technical scheme in the embodiment of the present invention is carried out to clear, complete description and discussion; obviously; as described herein is only a part of example of the present invention; it is not whole examples; based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under the prerequisite of not making creative work, belongs to protection scope of the present invention.
For the ease of the understanding to the embodiment of the present invention, be further explained as an example of several specific embodiments example below in conjunction with accompanying drawing, and each embodiment does not form the restriction to the embodiment of the present invention.
Embodiment mono-
The handling process of a kind of computing flow graph mapping method of optimizing many core system repairing performances that this embodiment provides as shown in Figure 2, comprises following treatment step:
Step 201, for a system computing flow graph having distributed, carries out Calculation of Reliability to the task of each node. Described computational methods, mainly consider importance and the probability of malfunction size of this task, certainly, also may consider some other factor. Task is important, and reliability ensures that priority is higher; Task probability of malfunction is higher, and reliability ensures that priority is higher. This step is the situation of each task of COMPREHENSIVE CALCULATING, and automatically provides reliability and ensure priority.
Step 202, the reliability of the each node task providing according to step 201 ensures priority, in conjunction with task node sum and the many core architecture resources of target situation, automatically or manually determine that each node need to reserve the number of contiguous idle core in mapping, make reliability and ensure priority list; Obviously, described many core architecture resources, are not limited only to the rectangular configuration of four quadruplications shown in Fig. 1, less more greatly rectangle, and circular or other similar structures are all included.
In the time that a task node reliability ensures that priority is higher, the idle check figure order of vicinity of distributing to this node is higher. Certainly, also can be arranged in a certain priority limit and give the idle core of certain vicinity reserved number.
A kind of reliability that this embodiment provides ensures priority list schematic diagram as shown in Figure 3. Each group packet of table need to be reserved the number of contiguous idle core during containing the title of node task and this node duty mapping. Meanwhile, owing to may can not meet the reserved requirement of certain some node in mapping, can ensure the idle check figure order of reserved vicinity in priority list according to corresponding situation amendment reliability.
The idle core of described vicinity, has different definition according to different situations; In the time that idle core and certain core distance is less than or equal to X, be the idle core of vicinity of this core, wherein X is more than or equal to 1 integer.
Step 203, the reliability providing based on step 202 ensures priority list, each node task of system is proceeded to from high to low one by one to the mapping of the many core frameworks of target according to priority.
This example provides a kind ofly ensures that according to system computing flow graph and node reliability steps flow chart that priority proceeds to the many core frameworks mappings of target configuration as shown in Figure 4, mainly comprises:
Step 400, initial step, in mapping method known system computing flow graph and computing flow graph, the reliability of each node ensures precedence information and the many core architecture resources of mapping target situation;
Step 401, in the node task of never mapping configuration, chooses reliability and ensures a highest node of priority, using this node as node to be mapped;
Step 402, checks in the many core frameworks of target whether also have unappropriated core, if also exist core not complete, carry out step 403, if all core is all assigned, carry out step 405;
Step 403, the reliability providing according to step 202 ensures priority list, check that the reliability that whether exists a panel region to meet this task node ensures requirement on many core frameworks, if cannot meet, carry out step 404, if can meet,, using this region as mapping target, carry out step 406;
Step 404, ensures in priority list in reliability, reduces this node mission reliability and ensures requirement, reduces the reserved contiguous idle check figure order of this node, re-executes step 403;
Step 405, because all nodes on many core frameworks are all assigned, choose from low to high according to reliability guarantee priority the task node having distributed, if this node has distributed reserved idle core, this mapping target using this reserved idle core as node to be allocated.
Step 406, step 403 and 405 has been determined the mapping target of task node to be allocated, node mapping to be mapped is arrived to target area, and the core of distribution is labeled as and is distributed;
Step 407, checks that whether all task nodes have all completed mapping, if do not completed, re-execute step 401;
Step 408, in range of distribution corresponding to each task node, as configuration core, configures corresponding task according to core of certain algorithm picks;
Step 409, mapping finishes, and completes many core system configuration, all core that do not configure is set to the free time, can operation core while breaking down as an alternative caryogamy put in the system of joining.
Such as, a kind of as shown in Fig. 5 Fig. 6 ensures that according to system computing flow graph and node reliability priority proceeds to the schematic diagram of the many core frameworks mappings of target configuration, wherein:
The each task node sidenote of system computing flow graph understands that the reliability of this node ensures priority, i.e. mapping configuration is to need the reserved idle check figure order of vicinity;
Ensure priority orders according to reliability, successively C, D, E, A, B, G and F node are distributed, it should be noted that in the time distributing F, there is no unallocated core, choose the reserved idle core of G as F mapping target core;
In each task node region, choose a core according to certain algorithm as configuration core, completion system configuration, its cokernel is as idle core to use when system selfreparing, and special, fault keranel is preferentially repaired in the region of distributing to this core.
The technical scheme being provided by the invention described above embodiment can be found out, the embodiment of the present invention is by carrying out Calculation of Reliability to each task node of system computing flow graph, for the reserved many idle core of the low node of reliability in its vicinity, break down when needing selfreparing for the low task core of reliability, can find in its vicinity fast idle core to complete selfreparing, efficiently completion system recovers fast; And because nuclear location is too far away apart from original position after avoiding occurring selfreparing, and need to move the situation of multiple core tasks, greatly reduce due to the impact of selfreparing on systematic function, especially for a fragile task node, repeatedly fault selfreparing is more effective.
Embodiment bis-
A kind of computing flow graph mapping device of optimizing many core system repairing performances that this embodiment provides, its concrete structure as shown in Figure 7, comprises as lower module:
Reliability prediction module 601, carries out Calculation of Reliability for each the node task to system computing flow graph, and generates reliability guarantee priority list;
Priority memory cell 602, ensures priority list for memory reliability;
Mapping dispensing unit 603, for ensureing that according to reliability priority list is mapped on the many core frameworks of target by system computing flow graph, completion system configuration;
Mapping area memory cell 604, for storing the mapping area of each task node.
Concrete, described reliability prediction module 601, according to the importance of this task and probability of malfunction size, determines the priority size of this task automatically, and task is important, and reliability ensures that priority is higher; Task probability of malfunction is higher, and reliability ensures that priority is higher. This two aspects factor need be considered, other factors can certainly be considered.
Concrete, described reliability prediction module 601, the priority list of generation is reserved contiguous idle check figure order with needs and is represented, and the idle check figure order of described vicinity can automatically generate or be configured by user.
Concrete, described priority memory cell 602, in the time that mapping distribution cannot complete, can modify, and reduces the idle check figure order of required reserved vicinity.
Concrete, described mapping dispensing unit 603, also, in the time that mapping distribution cannot complete, proposes amendment request to priority memory cell 602.
Concrete, described mapping area memory cell 604, also for when the system selfreparing, instructs the priority of task that needs to repair to be configured in the region that this core distributes.
The device of the application embodiment of the present invention completes that can to optimize mapping concrete steps and the preceding method embodiment of many core architecture system repairing performances similar, repeats no more herein.
One of ordinary skill in the art will appreciate that all or part of flow process in above-described embodiment method, can carry out the hardware that instruction is relevant by computer program to complete, described program can be stored in computer read/write memory medium, this program, in the time carrying out, can comprise the embodiment flow process of above-mentioned each method. Wherein, described storage medium is magnetic disc, CD, read-only store-memory body or random store-memory body etc.
In sum, the embodiment of the present invention is by carrying out Calculation of Reliability to each task node of system computing flow graph, for the reserved many idle core of the low node of reliability in its vicinity, break down when needing selfreparing for the low task core of reliability, can find in its vicinity fast idle core to complete selfreparing, efficiently completion system recovers fast; And because nuclear location is too far away apart from original position after avoiding occurring selfreparing, and need to move the situation of multiple core tasks, greatly reduce due to the impact of selfreparing on systematic function, especially for a fragile task node, repeatedly fault selfreparing is more effective.
The embodiment of the present invention can solve many core systems preferably because after fault, extend degradation problem under systematic function after selfreparing when selfreparing.
The above; only for preferably detailed description of the invention of the present invention, but protection scope of the present invention do not limit to therewith, is anyly familiar with in technical scope that those skilled in the art disclose in the present invention; the variation that can expect easily or replacement, within all should being encompassed in protection scope of the present invention. Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.
Claims (10)
1. a computing flow graph mapping method of optimizing many core system repairing performances, is characterized in that, comprising:
For the many core system computing flow graphs that distributed task, for the task of each node is enteredRow Calculation of Reliability, the reliability that generates each node task ensures priority;
Ensure that based on reliability priority generates the idle check figure order of the required reserved vicinity of node;
Ensure that based on reliability priority list and system computing flow graph complete computing flow graph to the many core framves of targetThe mapping of structure;
Ensure that according to reliability priority list and system computing flow graph complete computing flow graph to the many core framves of targetThe mapping of structure, comprises step:
The first step, chooses not reliability in mapping node and ensures the highest node of priority;
Whether bis-Walk, there is unallocated core on the many core frameworks of query aim, carries out tri-Walk if exist,Otherwise carry out Wu Walk;
The 3rd Walk, whether inquiry exists a panel region to meet the contiguous idle check figure requirement of node, if do not existCarry out the 4th step, otherwise carry out the 6th step;
The 4th step, reduces contiguous idle check figure requirement, re-starts the 3rd step;
The 5th step, choose reliability in mapping node ensure priority minimum and have contiguous idle coreThe node of number, using its idle core as mapping target;
The 6th step, arrives this region by node mapping;
The 7th step, it is complete whether inspection shines upon, and carries out the 8th step, otherwise get back to the first step if complete;
The 8th step selects a core to carry out as initial position according to certain method in each node regionConfiguration, completion system configuration;
The 9th step, all the other do not configure core as idle spare core.
2. the computing flow graph mapping method of the many core systems of optimization according to claim 1 repairing performance,It is characterized in that, described for the many core system computing flow graphs that distributed task, be eachThe task of node is carried out Calculation of Reliability, also comprises:
Task is important, and reliability ensures that priority is higher;
Task probability of malfunction is higher, and reliability ensures that priority is higher.
3. the computing flow graph mapping method of the many core systems of optimization according to claim 1 repairing performance,It is characterized in that, described according to the idle check figure of the reliability guarantee priority generation required vicinity taking of nodeOrder method, also comprises:
Generate contiguous idle core and need consider task node sum and many core architecture resources, different priorities distributesContiguous idle check figure order is generated or user's manual definition automatically by the method, is kept at reliability and ensures excellentIn first level list; Many core architecture resources are configured by user.
4. the computing flow graph mapping method of the many core systems of optimization according to claim 1 repairing performance,It is characterized in that, the scope of the idle core of described vicinity is defined by the user.
5. a computing flow graph mapping device of optimizing many core system repairing performances, is characterized in that, comprising:
Reliability prediction module, carries out reliability meter for each the node task to system computing flow graphCalculate, and generate reliability guarantee priority list;
Priority memory cell, ensures priority list for memory reliability;
Mapping dispensing unit, for ensureing that according to reliability priority list is mapped to by system computing flow graphOn the many core frameworks of target, completion system configuration;
Mapping area memory cell, for storing the mapping area of each task node.
6. the computing flow graph mapping device of the many core systems of optimization according to claim 5 repairing performance,It is characterized in that, described reliability prediction module, according to the importance of this task and probability of malfunction size,Automatically determine the priority size of this task;
Task is important, and reliability ensures that priority is higher; Task probability of malfunction is higher, and reliability is protectedBarrier priority is higher.
7. the computing flow graph mapping device of the many core systems of optimization according to claim 5 repairing performance,It is characterized in that, described reliability prediction module, the priority list of generation is reserved contiguous empty with needsNot busy check figure order represents, the idle check figure order of described vicinity is by the computing of the many core systems of described optimization repairing performanceFlow graph mapping device automatically generates or is configured by user.
8. the computing flow graph mapping device of the many core systems of optimization according to claim 5 repairing performance,It is characterized in that, described priority memory cell, in the time that mapping distribution cannot complete, amendment is to reduceThe idle check figure order of required reserved vicinity.
9. the computing flow graph mapping device of the many core systems of optimization according to claim 5 repairing performance,It is characterized in that,
Described mapping dispensing unit, also in the time that mapping distribution cannot complete, to priority memory cellAmendment request is proposed.
10. the computing flow graph mapping device of the many core systems of optimization according to claim 5 repairing performance,It is characterized in that described mapping area memory cell, also for when the system selfreparing, instructs and needs to repairPriority of task be configured in the mapping area of storing in mapping area memory cell.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310144403.9A CN103257904B (en) | 2013-04-24 | 2013-04-24 | Optimize computing flow graph mapping method and the device of many core system repairing performances |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310144403.9A CN103257904B (en) | 2013-04-24 | 2013-04-24 | Optimize computing flow graph mapping method and the device of many core system repairing performances |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103257904A CN103257904A (en) | 2013-08-21 |
| CN103257904B true CN103257904B (en) | 2016-05-04 |
Family
ID=48961838
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310144403.9A Active CN103257904B (en) | 2013-04-24 | 2013-04-24 | Optimize computing flow graph mapping method and the device of many core system repairing performances |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103257904B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP4120076A4 (en) * | 2020-04-01 | 2023-04-05 | Huawei Technologies Co., Ltd. | Task scheduling method and apparatus |
| CN113220548B (en) * | 2021-03-25 | 2024-02-09 | 中国航天系统科学与工程研究院 | Software reliability index distribution method, medium and equipment based on reliability block diagram |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556534A (en) * | 2009-04-21 | 2009-10-14 | 浪潮电子信息产业股份有限公司 | Large-scale data parallel computation method with many-core structure |
| CN102193830A (en) * | 2010-03-12 | 2011-09-21 | 复旦大学 | Many-core environment-oriented division mapping/reduction parallel programming model |
| US8327187B1 (en) * | 2009-09-21 | 2012-12-04 | Tilera Corporation | Low-overhead operating systems |
-
2013
- 2013-04-24 CN CN201310144403.9A patent/CN103257904B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101556534A (en) * | 2009-04-21 | 2009-10-14 | 浪潮电子信息产业股份有限公司 | Large-scale data parallel computation method with many-core structure |
| US8327187B1 (en) * | 2009-09-21 | 2012-12-04 | Tilera Corporation | Low-overhead operating systems |
| CN102193830A (en) * | 2010-03-12 | 2011-09-21 | 复旦大学 | Many-core environment-oriented division mapping/reduction parallel programming model |
Non-Patent Citations (4)
| Title |
|---|
| High Performance Fault-Tolerant Routing Algorithm for NoC-based Many-Core System;Masoumeh Ebrahimi等;《2013 21st Euromicro International Conference on Parallel,Distributed,and Network-Based Processing》;20130301;104-109 * |
| Parallax-A New Operating System Prototype Demonstrating Service Scaling and Service Self-Repair in Multi-core Servers;Dr.Rao Mikklineni等;《2011 20th IEEE International Workshops on Enabling Technologies:Infrastructure for Collaborative Enterprise》;20110629;462-469 * |
| 众核处理器系统核资源动态分组的自适应调度算法;曹仰杰等;《软件学报》;20120229;第23卷(第2期);240-252 * |
| 面向集成电路可靠性挑战的多核处理器虚拟化技术;张磊等;《信息技术快报》;20100331;第8卷(第2期);1-8 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103257904A (en) | 2013-08-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fizza et al. | Improving the schedulability of real-time tasks using fog computing | |
| Becchu et al. | System level fault-tolerance core mapping and FPGA-based verification of NoC | |
| KR20160106187A (en) | Prioritizing data reconstruction in distributed storage systems | |
| Li et al. | iSwitch: Coordinating and optimizing renewable energy powered server clusters | |
| CN102812439B (en) | For the method and system of assigned tasks in multiprocessor computer system | |
| US10664401B2 (en) | Method and system for managing buffer device in storage system | |
| CN102981910A (en) | Realization method and realization device for virtual machine scheduling | |
| US11144414B2 (en) | Method and apparatus for managing storage system | |
| CN101281493A (en) | And non flash memory device and management method thereof | |
| CN105988727B (en) | Storage method and storage device based on RAID | |
| KR20150016820A (en) | System and method for virtual machine placement and management on cluster system | |
| Khalili et al. | A fault-tolerant low-energy multi-application mapping onto NoC-based multiprocessors | |
| CN104753992A (en) | Method, device and system for data storage and method and device for virtual platform failure recovery | |
| CN102413186A (en) | Resource scheduling method and device based on private cloud computing, and cloud management server | |
| CN103513997A (en) | Program recording method and system used for bus network | |
| CN103257904B (en) | Optimize computing flow graph mapping method and the device of many core system repairing performances | |
| CN116954523B (en) | Storage system, data storage method, data reading method and storage medium | |
| Khalili et al. | A reliability-aware multi-application mapping technique in networks-on-chip | |
| US10394615B2 (en) | Information processing apparatus and job management method | |
| CN103294169B (en) | A kind of redundancy protection systems of many core systems of optimised power consumption and method | |
| CN104460938A (en) | System-wide power conservation method and system using memory cache | |
| WO2022038658A1 (en) | Task management device, task management method, and task management program | |
| US20210271400A1 (en) | Memory system and memory control method | |
| CN103714022A (en) | Mixed storage system based on data block | |
| CN104679603A (en) | Asymmetrical data triple modular redundancy method and asymmetrical data triple modular redundancy system |
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 |