CN102650950B - Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture - Google Patents
Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture Download PDFInfo
- Publication number
- CN102650950B CN102650950B CN201210102989.8A CN201210102989A CN102650950B CN 102650950 B CN102650950 B CN 102650950B CN 201210102989 A CN201210102989 A CN 201210102989A CN 102650950 B CN102650950 B CN 102650950B
- Authority
- CN
- China
- Prior art keywords
- gpu
- module
- virtual machine
- server
- client end
- 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
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000012545 processing Methods 0.000 title abstract description 9
- 230000005540 biological transmission Effects 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 claims description 8
- 238000005538 encapsulation Methods 0.000 claims description 6
- 230000004044 response Effects 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims description 3
- 230000001965 increasing effect Effects 0.000 abstract description 4
- 238000004364 calculation method Methods 0.000 abstract description 2
- 238000005265 energy consumption Methods 0.000 abstract description 2
- 230000007246 mechanism Effects 0.000 abstract description 2
- 230000007547 defect Effects 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 2
- HPTJABJPZMULFH-UHFFFAOYSA-N 12-[(Cyclohexylcarbamoyl)amino]dodecanoic acid Chemical compound OC(=O)CCCCCCCCCCCNC(=O)NC1CCCCC1 HPTJABJPZMULFH-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
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
- Computer And Data Communications (AREA)
Abstract
The invention provides a platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and a work method of the platform architecture. The platform architecture is used as a transmission medium by deploying middleware at a GPU server end and the end part of a virtual machine and using modes such as socket or infiniband to make up the defect that an original virtual machine platform cannot accelerate by using the GPU. The platform architecture can be used for managing the GPU resources through one or more centrally-controlled management nodes, carrying out fine grit division on the GPU resources and providing the function multitasking and execution. The virtual machine requests the GPU resources for the management nodes through the middleware and accelerates by using the GPU resources. The GPU server is used for registering the GPU resources for the management nodes through the middleware and providing service by using the GPU resources. According to the platform architecture disclosed by the invention, the parallel processing capability of the GPU is introduced into the virtual machine; and the utilization rate of the GPU is increased to the maximum extent by combining a management mechanism. According to the platform architecture, energy consumption can be effectively reduced and the calculation efficiency is increased.
Description
Technical field
The present invention relates to virtual machine in a kind of virtualized environment and utilize platform architecture and the method for work thereof of many GPU speed-up computation in redirected mode, belong to technical field of virtualization.
Background technology
Virtual be cloud computing core technology basis, the advantage such as cost savings, security enhancing that it brings obtains the accreditation of people gradually, is the study hotspot of computer science.Virtual by hardware resource of Intel Virtualization Technology; multiple identical computer hardware platforms can be simulated on one computer; thus multiple operating system can be run simultaneously and realize mutually isolated; improve the utilization ratio of server, have a large amount of application in server merging, network security, calculating data protection, high-performance calculation and the field such as credible.
In recent years, the performance of Graphics Processing Unit (GPU) and function significantly increased.The function of GPU is no longer confined to image procossing, and develop into one has the high processor calculating the highly-parallel of peak value and memory bandwidth simultaneously.Along with some support GPGPU(general graphical processing unit) release of technology (such as CUDA) that calculates, the application of GPGPU is also more and more extensive.Due to the powerful computation capability of GPU, make the heterogeneous schemas introducing CPU+GPU in increasing high performance computation.But the power consumption of GPU is comparatively large on the one hand, if each node is equipped with GPU, then greatly may increase the power consumption of cluster; On the one hand, the computation capability of GPU is powerful, and in most of computing, GPU is as coprocessor, and the parallel section only in speed code, makes the utilization rate of GPU not high; On the other hand due to the closure of GPU, virtual machine cannot directly utilize GPU to carry out speed-up computation.This makes the application of GPU in virtual machine be limited by very large.
Summary of the invention
technical matters
In order to solve the problem that virtualized environment cannot utilize GPU to accelerate, the present invention proposes a kind of many GPU vitualization platform architecture and the method for work thereof that are applicable to cluster, by Management Unit, client component, server-side component collaborative work, make virtual machine can obtain the powerful computation capability of GPU, the acquisition to GPU resource and fine-grained Resourse Distribute can be realized, and by carrying out load balancing to GPU, improve GPU utilization factor, reduce energy consumption.Present invention achieves the lifting utilizing GPU to virtual machine processing power, assembly in virtual machine is called GPU by interception application program, and be redirected to the machine privileged domain or long-range GPU server, the GPU of application program is invoked on privileged domain or long-range GPU server perform, and returns results to virtual machine complete.
technical scheme
The present invention adopts following technical scheme to solve the problems of the technologies described above:
A kind of platform architecture supporting many GPU vitualization, comprise GPU resource administration module, virtual machine client end module, GPU server module, GPU resource administration module is deployed on GPU resource management node, virtual machine client end module is deployed on virtual machine client end, GPU server module is deployed in GPU service end, GPU resource administration module is responsible for the registration of GPU server and the process to GPU resource request, virtual machine client end module and GPU server module carry out data transmission, virtual machine client end module and GPU server module are carried out alternately, virtual machine client end block intercepts virtual machine calling GPU, and be redirected to GPU server module, GPU server module then accepts the GPU recalls information of virtual machine client end block intercepts, call GPU to perform, and return execution result.
The method of work of the platform architecture of the many GPU vitualization of above-mentioned support comprises the steps:
Step 1, starts GPU resource administration module, monitors the registration request of GPU server module and the resource request of virtual machine client end module, and the GPU server maintenance state table to registration.
Step 2, starts GPU server module, sends registration request to GPU resource administration module.
Step 3, GPU resource administration module receives registration request, sets up a table, safeguards the state of current GPU server, return successfully after completing.
Step 4, GPU server module receives the information succeeded in registration, and monitors designated port at once.
Step 5, starts virtual machine client end module, monitors calling GPU.When virtual machine occurs that GPU calls, then send resource request to GPU resource administration module.
Step 6, GPU resource administration module receives resource request, obtains the duty of registered GPU server, according to certain algorithm, to the GPU server that virtual machine client end module assignment is mated most.
Step 7, GPU resource administration module receives the GPU server of distribution, and set up data with the GPU server module of this GPU server and transmit and be connected, virtual machine client end module calls encapsulation by interception, and is sent to GPU server module by data transmission connection.
Step 8, GPU server module receives the data of encapsulation, comprehensive each GPU load information, and be the GPU that its selection is mated most, execution is called, and returns results until perform end.
Step 9, virtual machine client end module receives result and returns to application program until perform end.
beneficial effect
The present invention utilizes existing Intel Virtualization Technology, the parallel processing capability of GPU is introduced virtual machine, and in conjunction with administrative mechanism, by the carrier that conventional socket, infiniband or the special communication mode of each virtual platform transmit as data, existing virtual platform is not modified, only on original platform base, add assembly, make virtual machine can use GPU to calculate, be applicable to all virtual platforms.The present invention easily uses, and user only needs simple setting and configuration, easily transplants, need not revise and can be operated on each virtual platform; Design Orientation of the present invention is in virtual cluster, GPU is utilized to assist to improve virtualized processing power, be applicable to teacher, student etc. and use GPU programming by demonstration technology or other guide at teaching process needs, also be adapted in virtual cluster simultaneously, improve the processing power of virtual machine, and improve the utilization rate of GPU.Use scenes of the present invention is extensive, has good practicality and feasibility.
Accompanying drawing explanation
Fig. 1 is high-level schematic functional block diagram of the present invention;
Fig. 2 is transmission procedure process flow diagram of the present invention;
Fig. 3 is real-time control routine process flow diagram of the present invention.
Embodiment
Below in conjunction with accompanying drawing, specific embodiments of the present invention are described:
As shown in Figure 1, the platform architecture of the many GPU vitualization of support provided by the invention comprises GPU resource administration module, virtual machine client end module, GPU server module.GPU resource administration module is deployed on GPU resource management node, and virtual machine client end module is deployed on virtual machine client end, and GPU server module is deployed in GPU service end (i.e. GPU server).Virtual machine client end is communicated by the exclusive mode of socket, infiniband or virtual platform with GPU service end and GPU resource administration module.GPU resource administration module is responsible for the registration of GPU server and the process to GPU resource request, virtual machine client end module and GPU server module carry out data transmission, virtual machine client end module and GPU server module are carried out alternately, virtual machine client end block intercepts virtual machine calling GPU, and be redirected to GPU server module, GPU server module then accepts the GPU recalls information of virtual machine client end block intercepts, calls GPU and performs, and return execution result.GPU resource administration module is to request place of GPU server module and virtual machine client end module
Reason, the resource registering of response GPU server module, and monitor the load of each GPU server in real time, the task of each GPU server is adjusted, the computational resource of each GPU server is managed, simultaneously according to the request of load response virtual machine client end module, and distribute for it GPU server mated most.Virtual machine client end module and GPU server module, undertaken alternately by socket or infiniband or the special communication mode of each virtual platform, and by interception, the mode that is redirected, the GPU completing virtual machine accelerates.In addition, GPU server module carries out resource management to GPU multiple in GPU server, and each GPU supports that multiple tasks in parallel is run, and GPU server module adds up each GPU current task load, and carries out the load balancing between GPU according to present load.
Be illustrated in figure 2 the workflow diagram of GPU resource management node.Virtual machine client end module is to GPU resource administration module request resource, and GPU server module is to GPU resource administration module registration resource.GPU resource administration module is resolved request, and responds according to current state.
Be illustrated in figure 3 virtual machine client end module and the mutual workflow diagram of GPU server module.Evaluation work completes jointly primarily of virtual machine client end module and GPU server module.After virtual machine client module and GPU server module connect, both sides carry out data transmit-receive transmission.Now, application program the calling for GPU that virtual machine client end block intercepts the machine is run, and send data to GPU server module.The data of GPU server module sink virtual machine client, and resolve, select GPU and utilize this GPU computing to obtain a result, and returning execution result.Virtual machine client end module receives data, and execution result is returned to application program.GPU server module judges whether the data that user side sends are stop transmission command.If not, then repeat said process; If so, then close this to connect.
Support that the method for work of the platform architecture of many GPU vitualization specifically comprises the steps:
Step 1, starts GPU resource administration module, monitors the registration request of GPU server module and the resource request of virtual machine client end module, and the GPU server maintenance state table to registration.
Step 2, starts GPU server module, sends registration request to GPU resource administration module.
Step 3, GPU resource administration module receives registration request, sets up a table, safeguards the state of current GPU server, return successfully after completing.
Step 4, GPU server module receives the information succeeded in registration, and monitors designated port at once.
Step 5, starts virtual machine client end module, monitors calling GPU.When virtual machine occurs that GPU calls, then send resource request to GPU resource administration module.
Step 6, GPU resource administration module receives resource request, obtains the duty of registered GPU server, according to certain algorithm, to the GPU server that virtual machine client end module assignment is mated most.
Step 7, GPU resource administration module receives the GPU server of distribution, and set up data with the GPU server module of this GPU server and transmit and be connected, virtual machine client end module calls encapsulation by interception, and is sent to GPU server module by data transmission connection.
Step 8, GPU server module receives the data of encapsulation, comprehensive each GPU load information, and be the GPU that its selection is mated most, execution is called, and returns results until perform end.
Step 9, virtual machine client end module receives result and returns to application program until perform end.
Claims (3)
1. support the method for work of the platform architecture of many GPU vitualization for one kind, it is characterized in that described platform architecture comprises GPU resource administration module, virtual machine client end module, GPU server module, GPU resource administration module is deployed on GPU resource management node, virtual machine client end module is deployed on virtual machine client end, GPU server module is deployed in GPU service end, GPU resource administration module is responsible for the registration of GPU server and the process to GPU resource request, virtual machine client end module and GPU server module carry out data transmission, virtual machine client end module and GPU server module are carried out alternately, virtual machine client end block intercepts virtual machine calling GPU, and be redirected to GPU server module, GPU server module then accepts the GPU recalls information of virtual machine client end block intercepts, call GPU to perform, and return execution result,
The request of described GPU resource administration module to GPU server module and virtual machine client end module processes, the registration request of response GPU server module, and monitor the load of each GPU server in real time, the task of each GPU server is adjusted, the computational resource of each GPU server is managed, simultaneously according to the request of load response virtual machine client end module, and distribute for it GPU server mated most;
Described method of work comprises following steps:
Step 1, starts GPU resource administration module, monitors the registration request of GPU server module and the resource request of virtual machine client end module, and the GPU server maintenance state table to registration;
Step 2, starts GPU server module, sends registration request to GPU resource administration module;
Step 3, GPU resource administration module receives registration request, sets up a state table, safeguards the duty of current GPU server, return the information succeeded in registration after completing;
Step 4, GPU server module receives the information succeeded in registration, and monitors designated port at once;
Step 5, starts virtual machine client end module, monitors calling GPU, when virtual machine occurs that GPU calls, then sends resource request to GPU resource administration module;
Step 6, GPU resource administration module receives resource request, obtains the duty of registered GPU server, according to certain algorithm, to the GPU server that virtual machine client end module assignment is mated most;
Step 7, GPU resource administration module receives the GPU server of distribution, and set up data with the GPU server module of this GPU server and transmit and be connected, virtual machine client end module calls encapsulation by interception, and is sent to GPU server module by data transmission connection;
Step 8, GPU server module receives the data of encapsulation, comprehensive each GPU load information, and be the GPU that its selection is mated most, execution is called, and returns results until perform end;
Step 9, virtual machine client end module receives result and returns to application program until perform end.
2. method of work according to claim 1, it is characterized in that described virtual machine client end module and GPU server module are undertaken alternately by socket or infiniband or the special communication mode of each virtual platform, by interception, redirected mode, the GPU completing virtual machine accelerates.
3. method of work according to claim 1, it is characterized in that described GPU server module carries out resource management to GPU multiple in GPU server, each GPU supports that multiple tasks in parallel is run, GPU server module adds up each GPU current task load, and carries out the load balancing between GPU according to current task load.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210102989.8A CN102650950B (en) | 2012-04-10 | 2012-04-10 | Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210102989.8A CN102650950B (en) | 2012-04-10 | 2012-04-10 | Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102650950A CN102650950A (en) | 2012-08-29 |
| CN102650950B true CN102650950B (en) | 2015-04-15 |
Family
ID=46692958
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210102989.8A Active CN102650950B (en) | 2012-04-10 | 2012-04-10 | Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102650950B (en) |
Families Citing this family (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103593246B (en) * | 2012-08-15 | 2017-07-11 | 中国电信股份有限公司 | Communication means, host and dummy machine system between virtual machine and host |
| CN103257885A (en) * | 2013-05-20 | 2013-08-21 | 深圳市京华科讯科技有限公司 | Media virtualization processing method |
| CN103257884A (en) * | 2013-05-20 | 2013-08-21 | 深圳市京华科讯科技有限公司 | Virtualization processing method for equipment |
| CN103309748B (en) * | 2013-06-19 | 2015-04-29 | 上海交通大学 | Adaptive scheduling host system and scheduling method of GPU virtual resources in cloud game |
| CN103324505B (en) * | 2013-06-24 | 2016-12-28 | 曙光信息产业(北京)有限公司 | The method disposing GPU development environment in group system and cloud computing system |
| CN103713938A (en) * | 2013-12-17 | 2014-04-09 | 江苏名通信息科技有限公司 | Multi-graphics-processing-unit (GPU) cooperative computing method based on Open MP under virtual environment |
| CN104754464A (en) * | 2013-12-31 | 2015-07-01 | 华为技术有限公司 | Audio playing method, terminal and system |
| CN105122210B (en) | 2013-12-31 | 2020-02-21 | 华为技术有限公司 | GPU virtualization implementation method and related device and system |
| US9584594B2 (en) | 2014-04-11 | 2017-02-28 | Maxeler Technologies Ltd. | Dynamic provisioning of processing resources in a virtualized computational architecture |
| US10715587B2 (en) | 2014-04-11 | 2020-07-14 | Maxeler Technologies Ltd. | System and method for load balancing computer resources |
| US9501325B2 (en) | 2014-04-11 | 2016-11-22 | Maxeler Technologies Ltd. | System and method for shared utilization of virtualized computing resources |
| CN105988874B (en) * | 2015-02-10 | 2020-08-28 | 阿里巴巴集团控股有限公司 | Resource processing method and device |
| WO2016145632A1 (en) * | 2015-03-18 | 2016-09-22 | Intel Corporation | Apparatus and method for software-agnostic multi-gpu processing |
| CN106155804A (en) * | 2015-04-12 | 2016-11-23 | 北京典赞科技有限公司 | Method and system to the unified management service of GPU cloud computing resources |
| CN105159753B (en) | 2015-09-25 | 2018-09-28 | 华为技术有限公司 | The method, apparatus and pooling of resources manager of accelerator virtualization |
| CN111865657B (en) * | 2015-09-28 | 2022-01-11 | 华为技术有限公司 | Acceleration management node, acceleration node, client and method |
| GB2545170B (en) * | 2015-12-02 | 2020-01-08 | Imagination Tech Ltd | GPU virtualisation |
| CN105528249B (en) * | 2015-12-06 | 2019-04-05 | 北京天云融创软件技术有限公司 | A kind of dispatching method of multiple users share GPU resource |
| CN105959404A (en) * | 2016-06-27 | 2016-09-21 | 江苏易乐网络科技有限公司 | GPU virtualization platform based on cloud computing |
| CN108121312B (en) * | 2017-11-29 | 2020-10-30 | 南瑞集团有限公司 | ARV load balancing system and method based on integrated hydropower management and control platform |
| US10846138B2 (en) | 2018-08-23 | 2020-11-24 | Hewlett Packard Enterprise Development Lp | Allocating resources of a memory fabric |
| CN109376011B (en) * | 2018-09-26 | 2021-01-15 | 郑州云海信息技术有限公司 | Method and device for managing resources in virtualization system |
| CN109388496A (en) * | 2018-11-01 | 2019-02-26 | 北京视甄智能科技有限公司 | A kind of image concurrent processing method, apparatus and system based on more GPU cards |
| CN109656714B (en) * | 2018-12-04 | 2022-10-28 | 成都雨云科技有限公司 | GPU resource scheduling method of virtualized graphics card |
| CN109582425B (en) * | 2018-12-04 | 2020-04-14 | 中山大学 | GPU service redirection system and method based on cloud and terminal GPU fusion |
| CN109598250B (en) * | 2018-12-10 | 2021-06-25 | 北京旷视科技有限公司 | Feature extraction method, device, electronic equipment and computer readable medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101938368A (en) * | 2009-06-30 | 2011-01-05 | 国际商业机器公司 | Virtual machine manager in blade server system and virtual machine processing method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8776050B2 (en) * | 2003-08-20 | 2014-07-08 | Oracle International Corporation | Distributed virtual machine monitor for managing multiple virtual resources across multiple physical nodes |
-
2012
- 2012-04-10 CN CN201210102989.8A patent/CN102650950B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101938368A (en) * | 2009-06-30 | 2011-01-05 | 国际商业机器公司 | Virtual machine manager in blade server system and virtual machine processing method |
Non-Patent Citations (2)
| Title |
|---|
| Distributed Processing, 2009. IPDPS 2009》.2009,第1-11页. * |
| Lin Shi等.vCUDA: GPU accelerated high performance computing in virtual machines.《IEEE International Symposium on Parallel & * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102650950A (en) | 2012-08-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102650950B (en) | Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture | |
| CN102929718B (en) | Distributed GPU (graphics processing unit) computer system based on task scheduling | |
| CN105912389B (en) | The virtual machine (vm) migration system under mixing cloud environment is realized based on data virtualization | |
| CN202565304U (en) | Distributed computing task scheduling and execution system | |
| CN102479100A (en) | Pervasive computing environment virtual machine platform and creation method thereof | |
| CN102271145A (en) | Virtual computer cluster and enforcement method thereof | |
| CN104615480A (en) | Virtual processor scheduling method based on NUMA high-performance network processor loads | |
| CN101860024B (en) | Implementation method for integrating provincial dispatch organization PAS system in electric power system | |
| CN104023062A (en) | Heterogeneous computing-oriented hardware architecture of distributed big data system | |
| Bala et al. | Offloading in cloud and fog hybrid infrastructure using iFogSim | |
| Montella et al. | Virtualizing high-end GPGPUs on ARM clusters for the next generation of high performance cloud computing | |
| CN103501295B (en) | A kind of remote access method based on virtual machine (vm) migration and equipment | |
| CN103685564A (en) | Plug-in application ability layer introduced industry application online operation cloud platform architecture | |
| CN102137162B (en) | CAD (Computer Aided Design) integrated system based on mode of software used as service | |
| CN105959404A (en) | GPU virtualization platform based on cloud computing | |
| Whaiduzzaman et al. | Pefc: Performance enhancement framework for cloudlet in mobile cloud computing | |
| Varghese et al. | Acceleration-as-a-service: Exploiting virtualised GPUs for a financial application | |
| Sarddar et al. | Central controller framework for mobile cloud computing | |
| CN104166581A (en) | Virtualization method for increment manufacturing device | |
| CN103747439A (en) | Wireless controller equipment, wireless authentication processing method, system and networking technique | |
| Chinenyeze et al. | An aspect oriented model for software energy efficiency in decentralised servers | |
| Liu et al. | BSPCloud: A hybrid distributed-memory and shared-memory programming model | |
| CN104699520B (en) | A kind of power-economizing method based on virtual machine (vm) migration scheduling | |
| CN104899094A (en) | Virtual machine load balance processing method | |
| Li et al. | Reliability evaluation for cloud computing system considering common cause failure |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20120829 Assignee: Jiangsu Wisedu Information Technology Co., Ltd. Assignor: Nanjing University of Aeronautics and Astronautics Contract record no.: 2013320000314 Denomination of invention: Platform architecture supporting multi-GPU (Graphics Processing Unit) virtualization and work method of platform architecture License type: Exclusive License Record date: 20130410 |
|
| LICC | Enforcement, change and cancellation of record of contracts on the licence for exploitation of a patent or utility model | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| EM01 | Change of recordation of patent licensing contract |
Change date: 20150421 Contract record no.: 2013320000314 Assignee after: JIANGSU WISEDU EDUCATION INFORMATION TECHNOLOGY CO., LTD. Assignee before: Jiangsu Wisedu Information Technology Co., Ltd. |
|
| LICC | Enforcement, change and cancellation of record of contracts on the licence for exploitation of a patent or utility model | ||
| EC01 | Cancellation of recordation of patent licensing contract |
Assignee: JIANGSU WISEDU EDUCATION INFORMATION TECHNOLOGY CO., LTD. Assignor: Nanjing University of Aeronautics and Astronautics Contract record no.: 2013320000314 Date of cancellation: 20150430 |
|
| LICC | Enforcement, change and cancellation of record of contracts on the licence for exploitation of a patent or utility model |