CN103136054A - Accelerating method based on open computing language - Google Patents
Accelerating method based on open computing language Download PDFInfo
- Publication number
- CN103136054A CN103136054A CN 201110375080 CN201110375080A CN103136054A CN 103136054 A CN103136054 A CN 103136054A CN 201110375080 CN201110375080 CN 201110375080 CN 201110375080 A CN201110375080 A CN 201110375080A CN 103136054 A CN103136054 A CN 103136054A
- Authority
- CN
- China
- Prior art keywords
- cloud computing
- available devices
- computing platform
- accelerated method
- application program
- 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.)
- Pending
Links
Landscapes
- Stored Programmes (AREA)
Abstract
The invention discloses an accelerating method based on an open computing language. The accelerating method comprises ensuring that application programs already access a plurality of available devices of a cloud computing platform, enabling the cloud computing platform to enquire properties of the plurality of devices; enabling the application programs to compare the properties of the plurality of devices with requirements of application; and enabling the application programs to select one of the plurality of devices on basis of a comparison result. Therefore, the accelerating method based on the open computing language utilizes local computation and cloud computation, enables proficiency of the cloud computing platform to be available for the application programs of the open computing language, the application programs do not need or rarely need to be suitable for use of cloud in ordinary or specified cloud implementation.
Description
Technical field
The present invention relates to a kind of accelerated method based on open computational language.
Background technology
Cloud computing (Cloud Computing) is the product that the traditional computer such as grid computing, Distributed Calculation, parallel computation, effectiveness calculating, the network storage, virtual, load balancing and network technical development merge.By making Computation distribution on a large amount of distributed computers, but not in local computer or remote server, the operation of enterprise data center will be more similar to the internet.This makes the enterprise can be with resource switch to the application of needs, according to demand access computer and storage system.
Open computational language is first opening, the royalty-free standard for the general multiple programming of heterogeneous system.Open computational language provides unified programmed environment for the software developer, so that for using multi-core CPU (CPU (central processing unit)), GPU (Graphics Processing Unit), cellular-type framework and writing efficient portable code such as high performance computing service device, desk side computer system and the handheld device of the various mixing of other parallel processors such as DSP (digital signal processor).This standard is by the exploitation of Khronos group.
In open computational language, endorse in parallel computer from main frame (normally CPU) and be unloaded to accelerator facility (for example, GPU, CPU or FPGA (field programmable gate array)) same system.In addition, open computational language clearly cover mobile and embedded device so that develop the portable computing intensive application.Yet in foreseeable future, the computation capability of mobile device may limited.Although this may be enough for little low latency figure workload, attempt the open computational language application of operation computation-intensive (such as the simulation in science, engineering and business calculating, complex data analysis etc.) and will cause disappointed user to experience.In addition, probably will there be ultra-light-weight or the embedded platform that does not include the equipment of open computational language ability fully and have the very limited CPU of performance.Complicated open computational language is used and will can not be moved in these systems fully.
Even on standard desktop machine and workstation, the open computational language of computation-intensive is used and can is unloaded to the server farm that calculates in cloud and accelerate by opening the computational language workload.Yet, make the existing interface of operation workload in cloud may need using the remarkable modification of itself.These modifications might be limited to the particular cloud computing system, and this has even further hindered and adopt cloud computing in industry.
Summary of the invention
For the above-mentioned deficiency of prior art, the present invention proposes a kind of accelerated method based on open computational language, comprising:
Determine that application program accessed a plurality of available devices of a cloud computing platform;
The attribute of the described a plurality of available devices of described cloud computing platform inquiry;
Described application program compares the attribute of described a plurality of available devices and the requirement of application; And
Based on the result of described contrast, described application program is selected one of described a plurality of available devices.
According to a preferred embodiment of the present invention, in above-mentioned accelerated method, after described application program is selected one of described a plurality of available devices, described accelerated method further comprises: described application program is invoked at the brief introduction of the upper establishment of one of these a plurality of available devices document via one, and wherein said document brief introduction is used for doing further mutual with one of these a plurality of available devices.
According to a preferred embodiment of the present invention, in above-mentioned accelerated method, the step of the attribute of the described a plurality of available devices of described cloud computing platform inquiry further comprises: described cloud computing platform calls in response to one at least one virtual unit is added one or more as in described a plurality of available devices.
According to a preferred embodiment of the present invention, in above-mentioned accelerated method, described virtual unit represents the available resources in described cloud computing platform, and the hardware characteristics of the attribute description corresponding system of described virtual unit.
According to a preferred embodiment of the present invention, in above-mentioned accelerated method, described a plurality of available devices comprise processor, and described cloud computing platform is coupled to described processor via a network.
According to a preferred embodiment of the present invention, in above-mentioned accelerated method, described virtual unit represents in described cloud computing platform same type and has one group of isomery physical system of same alike result.
Therefore, the present invention utilizes local computing and cloud computing, makes the potential of cloud computing platform become and can use open computational language application program, and seldom need or do not need to make these application programs to adapt in general or particular cloud realize to use cloud.In addition, " enabling cloud " open computational language framework can allow available computing power on open computational language applications exploiting server platform, thereby causes experiencing across super function and/or the user of the client form factor of wide scope.
Should be appreciated that the above generality of the present invention is described and the following detailed description is all exemplary and explanat, and be intended to the further explanation that the invention provides for as claimed in claim.
Description of drawings
Accompanying drawing is mainly be used to providing, the present invention further to be understood.Accompanying drawing shows embodiments of the invention, and plays the effect of explaining the principle of the invention together with this instructions.In accompanying drawing:
Fig. 1 schematically shows according to the accelerated method based on open computational language of the present invention.
Embodiment
Describe technical scheme of the present invention in detail below in conjunction with accompanying drawing.
Fig. 1 schematically shows according to the accelerated method based on open computational language of the present invention.In the embodiment shown in fig. 1, the accelerated method 100 based on open computational language of the present invention comprises:
Step 101: determine that application program accessed a plurality of available devices of a cloud computing platform;
Step 102: the attribute of the described a plurality of available devices of described cloud computing platform inquiry;
Step 103: described application program compares the attribute of described a plurality of available devices and the requirement of application; And
Step 104: based on the result of described contrast, described application program is selected one of described a plurality of available devices.
Preferably, after described application program is selected one of described a plurality of available devices, described accelerated method further comprises: described application program is invoked at the brief introduction of the upper establishment of one of these a plurality of available devices document via one, and wherein said document brief introduction is used for doing further mutual with one of these a plurality of available devices.
In addition, preferably, the step of the attribute of the described a plurality of available devices of described cloud computing platform inquiry further comprises: described cloud computing platform calls in response to one at least one virtual unit is added one or more as in described a plurality of available devices.According to a preferred embodiment of the present invention, in above-mentioned accelerated method, described virtual unit represents the available resources in described cloud computing platform, and the hardware characteristics of the attribute description corresponding system of described virtual unit.
In better accelerated method, described a plurality of available devices comprise processor, and described cloud computing platform is coupled to described processor via a network.
According to a preferred embodiment of the present invention, in above-mentioned accelerated method, described virtual unit represents in described cloud computing platform same type and has one group of isomery physical system of same alike result.
The document brief introduction can be used for doing further mutual with this equipment subsequently in follow-up step.In one embodiment, a cloud strengthens driver and for example virtual unit is added in response to calling clGetDeviceIds () the available devices list of returning.This virtual unit represents the available resources in cloud, and the hardware characteristics of its attribute description corresponding system.
In certain embodiments, cloud computing platform is made of the server farm with powerful and/or multi-core CPU, so the attribute of virtual unit " CL_ equipment _ type " will be set as " CL_ equipment _ type _ CPU ".Yet cloud system can comprise GPU (Graphics Processing Unit), accelerator etc., and in this case, device type will be respectively " CL_ equipment _ type _ GPU " or " CL_ equipment _ type _ accelerator ".This means that each virtual unit can represent in cloud computing platform same type and have one group of isomery physical system of same alike result.In certain embodiments, cloud computing platform can be by realizing the virtual unit of Type C L_DEVICE_TYPE_CPU at the identical virtual machine of isomery physical system deploy.Therefore, in fact the attribute of virtual unit will reflect the configuration of the virtual machine on the physical system that will be deployed in cloud.In order to use virtual unit, application will from list select this equipment and each identical open computational language function used as local device.Therefore, application can be by inquiring about that these attributes are determined on cloud computing system or whether meaningful at the given open computational language kernel of local runtime.In certain embodiments, application code does not need modification just can utilize cloud computing system.On the contrary, cloud computing system can seamlessly be incorporated in open computational language framework and by using only and be selected based on its open computational language attribute.
Therefore, some embodiment utilize local computing and cloud computing.For example, the Resource Abstract/management that is provided by cloud computing and data transmission capabilities and agreement (such as the web/ network service) can be utilized and be incorporated in open computational language framework via virtual open computational language equipment.Therefore, it is available to open computational language application that the potential of cloud computing platform becomes, and seldom need or do not need to make these application adaptation to use cloud in general or particular cloud realize.And, with the interface of cloud computing platform can be encapsulated in alternately in virtual open computational language equipment and by open computational language driver handles.In addition, " enabling cloud " open computational language framework can allow available computing power on open computational language applications exploiting server platform, thereby causes experiencing across super function and/or the user of the client form factor of wide scope.For example, the ability that is usually provided by server farm that comprises can be provided the computing power of thin equipment.In addition, can be used as new commerce services open computational language cloud service is provided, for example, open computational language driver can provide free and charge by using.
Above-described embodiment is to provide to those of ordinary skills and realizes or use of the present invention; those of ordinary skills can be in the situation that do not break away from invention thought of the present invention; above-described embodiment is made various modifications or variation; thereby protection scope of the present invention do not limit by above-described embodiment, and should be the maximum magnitude that meets the inventive features that claims mention.
Claims (6)
1. accelerated method based on open computational language comprises:
Determine that application program accessed a plurality of available devices of a cloud computing platform;
The attribute of the described a plurality of available devices of described cloud computing platform inquiry;
Described application program compares the attribute of described a plurality of available devices and the requirement of application; And
Based on the result of described contrast, described application program is selected one of described a plurality of available devices.
2. accelerated method as claimed in claim 1, is characterized in that, after described application program was selected one of described a plurality of available devices, described accelerated method further comprised:
Described application program is invoked at the brief introduction of the upper establishment of one of these a plurality of available devices document via one, and wherein said document brief introduction is used for doing further mutual with one of these a plurality of available devices.
3. accelerated method as claimed in claim 1, is characterized in that, the step of the attribute of the described a plurality of available devices of described cloud computing platform inquiry further comprises:
Described cloud computing platform calls in response to one at least one virtual unit is added one or more as in described a plurality of available devices.
4. accelerated method as claimed in claim 3, is characterized in that, described virtual unit represents the available resources in described cloud computing platform, and the hardware characteristics of the attribute description corresponding system of described virtual unit.
5. accelerated method as claimed in claim 1, is characterized in that, described a plurality of available devices comprise processor, and described cloud computing platform is coupled to described processor via a network.
6. accelerated method as claimed in claim 3, is characterized in that, described virtual unit represents in described cloud computing platform same type and has one group of isomery physical system of same alike result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110375080 CN103136054A (en) | 2011-11-22 | 2011-11-22 | Accelerating method based on open computing language |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110375080 CN103136054A (en) | 2011-11-22 | 2011-11-22 | Accelerating method based on open computing language |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103136054A true CN103136054A (en) | 2013-06-05 |
Family
ID=48495912
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110375080 Pending CN103136054A (en) | 2011-11-22 | 2011-11-22 | Accelerating method based on open computing language |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103136054A (en) |
-
2011
- 2011-11-22 CN CN 201110375080 patent/CN103136054A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Moges et al. | Energy-aware VM placement algorithms for the OpenStack Neat consolidation framework | |
| Varghese et al. | Challenges and opportunities in edge computing | |
| US20190356734A1 (en) | Optimization of integration flows in cloud environments | |
| Calheiros et al. | CloudSim: a toolkit for modeling and simulation of cloud computing environments and evaluation of resource provisioning algorithms | |
| US9367359B2 (en) | Optimized resource management for map/reduce computing | |
| US8843621B2 (en) | Event prediction and preemptive action identification in a networked computing environment | |
| Zhang et al. | Container-VM-PM architecture: A novel architecture for docker container placement | |
| Singh et al. | Agent based framework for scalability in cloud computing | |
| Kim et al. | Enabling Digital Earth simulation models using cloud computing or grid computing–two approaches supporting high-performance GIS simulation frameworks | |
| Montella et al. | Virtualizing high-end GPGPUs on ARM clusters for the next generation of high performance cloud computing | |
| Dubey et al. | Efficient VM placement policy for data centre in cloud environment | |
| Ibrahim et al. | Resource allocation algorithm for GPUs in a private cloud | |
| Sanabria et al. | New heuristics for scheduling and distributing jobs under hybrid dew computing environments | |
| CN115361285B (en) | Method, device, equipment and medium for realizing off-line service mixed deployment | |
| Park et al. | Cloud computing platform for GIS image processing in U-city | |
| Khattar et al. | Multi-criteria-based energy-efficient framework for VM placement in cloud data centers | |
| Rausch et al. | A system for operating energy-aware cloudlets | |
| Mehta et al. | Design of infrastructure as a service (IAAS) framework with report generation mechanism | |
| Giannoutakis et al. | Evaluation of self‐organizing and self‐managing heterogeneous high performance computing clouds through discrete‐time simulation | |
| Pattnaik et al. | Load balancing in cloud computing environment using CloudSim | |
| CN103136054A (en) | Accelerating method based on open computing language | |
| Guo et al. | Optimized dependent file fetch middleware in transparent computing platform | |
| Vivek et al. | Payload fragmentation framework for high-performance computing in cloud environment | |
| Singh et al. | Numerical simulation design of multiple users offloading using improved optimization approach for edge computing | |
| CN112148935A (en) | Method and apparatus for NBMP function execution for multiple instances |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130605 |