CN105389220B - A Method of Parallel Linear Algebra Computation in Interactive R Language Platform - Google Patents
A Method of Parallel Linear Algebra Computation in Interactive R Language Platform Download PDFInfo
- Publication number
- CN105389220B CN105389220B CN201510755923.2A CN201510755923A CN105389220B CN 105389220 B CN105389220 B CN 105389220B CN 201510755923 A CN201510755923 A CN 201510755923A CN 105389220 B CN105389220 B CN 105389220B
- Authority
- CN
- China
- Prior art keywords
- platform
- interactive
- language
- linear algebra
- parallel linear
- 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
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 91
- 238000000034 method Methods 0.000 title claims abstract description 30
- 239000011159 matrix material Substances 0.000 claims abstract description 76
- 230000006870 function Effects 0.000 claims abstract description 37
- 238000013461 design Methods 0.000 claims abstract description 6
- 230000007812 deficiency Effects 0.000 claims abstract description 3
- 238000004364 calculation method Methods 0.000 claims description 20
- 238000005516 engineering process Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 5
- 238000011161 development Methods 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 238000005316 response function Methods 0.000 claims 3
- 230000007613 environmental effect Effects 0.000 abstract description 7
- 230000004044 response Effects 0.000 abstract description 2
- 238000011156 evaluation Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/54—Interprogram communication
- G06F9/547—Remote procedure calls [RPC]; Web services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5011—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resources being hardware resources other than CPUs, Servers and Terminals
- G06F9/5022—Mechanisms to release resources
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/54—Indexing scheme relating to G06F9/54
- G06F2209/541—Client-server
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Devices For Executing Special Programs (AREA)
Abstract
The method for the parallelization linear algebra calculating based on interactive R language platform that the invention discloses a kind of, the following steps are included: providing two computing platforms, one is interactive R language platform, another is parallel linear algebra computing platform, and two computing platforms are communicated by computer network；Then in interactive R language platform, the application programming interfaces that a parallel linear algebra calculates are realized in design；Finally in the distributed matrix class for the application programming interfaces that parallel linear algebra calculates, member variable comprising a R environmental form, and in the initialization procedure of the object in distributed matrix class, the garbage reclamation response mode of the member variable is registered to the Garbage Collector of interactive R language platform by the reg.finalizer function of R language.The present invention, which solves existing interactive mode R Programming with Pascal Language platform, can not carry out the deficiency of parallel linear algebra calculating, extend the computing capability of interactive R language platform.
Description
Technical field
The present invention relates to parallel computing more particularly to it is a kind of can be in interactive (interactive) R language platform
The method for carrying out parallel linear algebra calculating.
Background technique
R language is widely used a kind of programming language in data science field.R language provides for computer user
A large amount of common statistics calculate functions, and support the user oneself to write program and be extended to the function of R language.R language
It itself provides batch processing function library, they constitute original R language platform.User oneself can write program to R language
Speech is extended, and the program that user writes generally is extended in R language platform in the form of R lingware packet.R language is with function
Formula is programmed for main programming paradigm, while the modern procedures design method such as support Object-oriented Programming Design.
R language platform supports batch processing (batch) operation and interactive (interactive) to run two kinds of operation sides simultaneously
Formula.The interactive method of operation provides an interactive order line console for computer user, and user can be alternatively to R
Instruction is inputted in platform, R language platform calculate according to instruction and carries out response to user.The interactive method of operation allows
User on one side design, while adjust, fed back near real-time, mistake and shortcoming made to be corrected and be supplemented in time, side
R LISP program LISP development process, by the welcome of data science man.According to O ' Reilly publishing house of U.S. sum
The investigation of the website KDNuggets, R language are widely used in data science person's community.Since R language platform is initially for list
Computer environment design, therefore basic R language platform itself is unable to fully utilize multiple processors (CPU) or more meters
The computation capability that calculation machine provides.In the big data era that data scale is increasing, basic R language platform is limited to list
The computing capability of computer and large-scale data analysis task can not be handled.Basic R language platform processing capacity it is limited one
A main performance is exactly: basic R language platform can not handle large-scale linear algebra computational problem (such as extensive matrix
Multiplication).Extensive linear algebra calculating can be carried out by how extending basic R language platform, it has also become computing technique field
One major issue for needing to solve.And extensive linear algebra calculates the means solution mainly by parallel computing at present
Certainly.
On the other hand, the existing parallel computing based on message passing interface (MPI), is able to solve extensive line
Property algebraic manipulation problem.Software library ScaLAPACK based on message passing interface (MPI) technology provides one group of application program and connects
Mouth (API) function covers most linear algebra and calculates demand.ScaLAPACK software library is by using MPI parallel computation
Technology breaches the limitation of unicomputer computing capability, can make full use of the computing capability of multiple stage computers.It is existing at present
PbdR project (network address of the project is http://r-pbd.org/) carries out function to R language platform using ScaLAPACK software library
It can extend, R language users is allow to carry out parallel linear algebra calculating.But write based on ScaLAPACK software library or
Program based on pbdR item development can only be run in a manner of batch processing, and it is flat that they can not directly run on interactive R language
In platform.This limitation is so that the user of interactive mode R language platform can not be carried out on a large scale using the library ScaLAPACK, pbdR project
Parallel linear algebra calculate.It there is no method that can carry out parallel linear algebra calculating in interactive R language platform at present.
Summary of the invention
Goal of the invention: parallel linear algebra calculating can not be carried out not to overcome in existing interactive mode R language platform
Foot, the present invention provides a kind of method for extending interactive mode R language platform, this method keeps user flat in interactive R language
Parallel linear algebra calculating is carried out in platform, the specific implementation details calculated without understanding parallel linear algebra solves existing
Interactive R Programming with Pascal Language platform can not carry out the deficiency of parallel linear algebra calculating, extend the computing capability of interactive R.
The technical solution of the invention is as follows: in order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention one
Kind is based on the calculation method of client-server (Client-Server) model, and this method is by interactive computing platform and parallel
Linear algebra computing platform mutually separates in structure, and the cooperated computing of two platforms is realized by computer network communication.Entirely
The technical scheme comprises the following steps:
(1) method of the invention provides two computing platforms, and one is interactive R language platform, another is parallel
Linear algebra computing platform, two computing platforms are communicated by computer network, cooperated computing；
(2) in interactive R language platform, a distributed matrix class is defined, such, which is provided, interactive to run
Parallel linear algebra calculate application programming interfaces；
It (3) include the member variable of R environment (environment) type in distributed matrix class；
(4) garbage reclamation receptance function is registered for the member variable of R environmental form in distributed matrix class.
Further, two computing platforms in the step (1) are respectively: one, interactive mode R language platform, this is flat
Platform is the interactive R language platform of a standard, and interactive mode R language platform is loaded with the extended software packet that the present invention realizes,
And it is directly interacted with computer user；Two, parallel linear algebra computing platform, the computing platform are one based on MPI
(message passing interface) technology and ScaLAPACK software library, realize the computing platform of parallel linear algebra computing function.Two
A computing platform is communicated with each other by computer network.Interactive R language platform receives interactive calculating from computer user
Order, and corresponding computations are sent to parallel linear algebra computing platform, the latter carries out specific parallel linear algebra meter
It calculates and calculated result is returned into interactive R language platform, and result is fed back into computer by interactive R language platform and is used
Family.
Further, in the step (2) " distributed matrix class " refer to a R language S3 class either S4 class.
Such is write using R language, and is loaded into interactive R language platform, for users to use.The distributed matrix class provides
One group of R language function calculated for parallel linear algebra, user are corresponding by calling on the object of the distributed matrix class
Calculating function, to complete parallel linear algebra calculating task.
Further, indicate that " distributed matrix class " should include a special member variable in the step (3),
The type of the member variable is R environment (environment) type.The major function of the member variable is: when interactive R language
When Garbage Collector in platform recycles the object of distributed matrix class, Garbage Collector is registered by the member variable
Receptance function, notify parallel linear algebra computing platform, deletion is synchronously carried out to corresponding matrix data.
Further, in the step (4), the process of registration garbage reclamation receptance function is the structure in distributed matrix class
It makes and is completed in function, realized by the reg.finalizer function of R language itself.
The beneficial effects of the present invention are: (1) is by mutually dividing interactive R language platform with parallel linear algebra computing platform
From both making with the different methods of operation while to run.By computer network communication, enable interactive R language platform
The function of calling parallel linear algebra computing platform to provide, without losing its interactivity.The present invention solves cannot be in interactive R
The problem of parallel linear algebra calculating is carried out in language platform.(2) by the way that all parallel linear algebras calculating function wrapping exists
In one distributed matrix class, the present invention provided in interactive R language platform one it is user-friendly using interface.Meter
Calculation machine user can be as operation with traditional R language single machine matrix general operation distributed matrix, to carry out parallel linear algebra calculating.
The R language calculation code that the present invention writes keeps height consistent with primary R language calculation code, and alleviating computer user makes
Learning cost when with the present invention.(3) present invention registers rubbish to Garbage Collector by the reg.finalizer function of R language
Rubbish recycles receptance function, and the rubbish of the garbage reclamation and parallel linear algebra computing platform that solve interactive R language platform returns
The stationary problem of receipts.The present invention technical solution that uses, which avoids, simultaneously in similar technique needs that C language is used to be programmed
Problem allows the present invention all with R Programming with Pascal Language realization, to reduce the implementation complexity of entire technical solution.(4) make
After extending interactive mode R language platform with the present invention, interactive R language platform can break through the processing capacity limit of unicomputer
System, and carry out more massive linear algebra calculating.
Detailed description of the invention
Fig. 1 is disposed of in its entirety flow diagram of the invention.
Fig. 2 is the performance comparison figure for the single machine linear algebra computing system that the present invention is provided with standard R language platform.
Specific embodiment
In the following with reference to the drawings and specific embodiments, the present invention is furture elucidated, it should be understood that these embodiments are merely to illustrate
It the present invention rather than limits the scope of the invention, after the present invention has been read, those skilled in the art are to of the invention each
The modification of kind equivalent form falls within the application range as defined in the appended claims.
Technical solution of the present invention is mainly made of two software modules: one is interactive R language platform, in addition one
A is parallel linear algebra computing platform.Interactive R language platform is the R language computing platform an of standard, from R
The realization (project network address is https: //www.r-project.org/) of Project project, this software is not belonging to the present invention
Content.And parallel linear algebra computing platform is then the computing platform for meeting following feature: (1) computing platform can lead to
It crosses computer network and receives computations, and can analyze the instruction, execute corresponding parallel linear algebra calculating task；(2) should
Platform can storage matrix data in a distributed manner；(3) computing platform pass through MPI(message passing interface) technology and
ScaLAPACK software library progress parallel linear algebra calculating (MPI technology and ScaLAPACK software library are not belonging to the content of present invention,
Mentioned in description of the invention " ScaLAPACK software library " generation refer to that " compatible with ScaLAPACK application programming interfaces standard is soft
Part library ")；(4) identifier of calculated result or calculated result can be returned to calculating by computer network by the computing platform
Instruct sender.The computing platform for all meeting above-mentioned 4 features can be considered as meeting the parallel of the technology of the present invention feature
Linear algebra computing platform.
User is as shown in Figure 1 using the detailed process that the present invention carries out parallel linear algebra calculating.Interactive R language is flat
Platform is directly interacted with final computer user, is received the instruction that user gives, is executed calculating, and calculated result is passed through
Computer screen returns to user.For the parallel linear algebra computations that user assigns, interactive R language platform passes through meter
The instruction is sent to parallel linear algebra computing platform, is passed through by parallel linear algebra computing platform by calculation machine network
ScaLAPACK software library executes calculating operation, and calculated result is returned to interactive R language platform by computer network,
Computer user is finally returned result to by interactive R language platform.
The initialization Booting sequences of two computing platforms of the present invention the following steps are included:
(1) user starts an interactive R language platform on their computer；
(2) user loads the R software package realized by technical solution of the present invention in interactive R language platform, and to interaction
Formula R language platform issues the instruction that starting parallel linear algebra calculates；
(3) interactive mode R language platform passes through the Operation control mechanism of MPI technology, starts parallel linear algebra computing platform
Corresponding MPI operation；
(4) after the starting of parallel linear algebra computing platform, interactive R language platform and parallel linear algebra computing platform
Master computing node establishes computer network communication link；
(5) interactive mode R language platform creates an empty queue as global garbage reclamation queue, then computer is notified to use
Family software language platform loading of the present invention finishes, and interactive R language platform enters waiting instruction phase；
(6) user executes parallel linear algebra calculating task by interactive mode R language platform.
User is not directly facing the library ScaLAPACK and MPI technology is programmed use in interactive R language platform, and
It is to be programmed realization towards distributed matrix class provided by the invention.The embodiment of distributed matrix class of the invention are as follows:
It is then such by way of programming firstly, defining the distributed matrix class using a S3 class or S4 class in R language
Specific member variable is provided；Such member variable includes: (1) long-range matrix identifier variable, and the identifier is in R language
Type can be integer or character string, as long as different long-range matrixes, (2) R environment can be distinguished
(environment) variable of type, the variable are used for the garbage reclamation of distributed matrix object；Then, by R language towards
Object program design mechanism provides several member functions for such, and the existing linear algebra of heavy duty R language calculates function, makes
These calculate function and support distributed matrix class.
It is calculated in function in the member function either linear algebra of each distributed matrix class, completes specific parallel line
The step of property algebraic manipulation task are as follows:
(1) interactive mode R language platform reads the remote of the distributed matrix class object for participating in calculating from function call parameter
Journey matrix identifier；
(2) interactive mode R language platform counts this corresponding computations of linear algebra calculating operation and all participations
The long-range matrix identifier of calculation, global garbage reclamation queue, pass to parallel linear algebra computing platform by computer network；
(3) parallel linear algebra computing platform receives corresponding computations and participates in the long-range matrix identification calculated
Symbol, global garbage reclamation queue.Parallel linear algebra computing platform then reads out its preservation from global garbage reclamation queue
The long-range matrix identifier of rubbish, the corresponding matrix function of the long-range matrix identifier of these rubbish is then deleted from computing platform
According to.Then analytical Calculation instructs parallel linear algebra computing platform, and according to the long-range matrix identifier for participating in calculating, from memory
It is middle to take out the distributed matrix operation handle being stored in the computing platform；
(4) parallel linear algebra computing platform by the distributed matrix taken out in previous step operate handle and
The library ScaLAPACK carries out parallel linear algebra calculating；
(5) parallel linear algebra computing platform save this calculate generate as a result, the result be a distributed matrix,
It and is the long-range matrix identifier of distribution of results formula matrix allocation；
(6) the long-range matrix identifier that parallel linear algebra computing platform generates previous step is packaged into replying instruction,
Interactive R language platform is sent to by computer network；
(7) interactive mode R language platform receives the replying instruction that parallel linear algebra computing platform is sent back to, and is taken out long-range
Matrix identifier, the object of a new distributed matrix class is initialized using the identifier, and the object is returned to user.
The object of the distributed matrix class of return can be used in user, carries out other calculating tasks.
When using R language definition distributed matrix class, it is desirable to provide an initialization constructed fuction makes interactive R language
Speech platform can correctly handle the garbage reclamation problem of the object of distributed matrix class.The initialization constructed fuction of distributed matrix class
Workflow comprise the steps of:
(1) in calculator memory, initialization constructed fuction receives a long-range matrix identifier as the defeated of the function
Enter parameter；
(2) initialization constructed fuction creates an empty distributed matrix object in memory；
(3) it is created in the long-range matrix identifier initialization step (1) in initialization constructed fuction utilization input parameter
The long-range matrix identifier member variable of distributed matrix object；
(4) distributed matrix Object Creation one new R environment (environment) type to be created in step (1)
Member variable, and by input parameter in long-range matrix identifier be stored in the member variable of the R environmental form；
(5) the reg.finalizer function for passing through R language, R environmental form member variable described in (4) step is existed
It is registered in the Garbage Collector of interactive R language platform, by the R environmental form member variable and a customized rubbish
Recycling receptance function is associated；
(6) by distributed matrix object creating in step (1), being handled by above-mentioned steps again, as construction letter
Several return values returns.
Further, (5) step in the workflow of the initialization constructed fuction of above-mentioned distributed matrix class, is mentioned
One customized garbage reclamation receptance function, the workflow of the function comprise the steps of:
(1) customized garbage reclamation receptance function receives a R environmental form variable as input parameter；
(2) receptance function reads long-range matrix identifier from R environmental form variable, which is added to global rubbish
Rubbish recycles in queue.
The invention has the advantages that computer user is allow to carry out parallel linear generation in interactive R language platform
Number calculates；By way of multiple stage computers parallel computation, the present invention can allow user to handle in interactive R language platform greatly
When the linear algebraic manipulation problem of scale, obtains and calculate faster calculating speed than single machine linear algebra.The present invention is based on existing
Some open source softwares realize a prototype system.It requires according to the technique and scheme of the present invention, prototype system includes two calculating
Platform, wherein the interactive R language platform that interactive mode R language platform uses R Project project to provide, and parallel linear algebra
Computing platform is then the prototype computing platform developed according to the technique and scheme of the present invention, has used pbdR project in exploitation
The software that (project home page http://r-pbd.org/) is provided.The software that R Project project and pbdR project provide is not belonging to
The contents of the present invention.According further to technical solution of the present invention, in prototype system further including one, to run on interactive R language flat
The distributed matrix class of platform.Benchmark test is used as by using matrix multiplication operation (a kind of linear algebra calculating operation), to this
The single machine linear algebra computing system that the prototype software system realized and existing R language platform provide is invented to be tested,
Evaluation and test uses calculating time-consuming as Measure Indexes.In evaluation and test, the prototype software system that the present invention realizes has used 10 computers
Parallel computation is carried out, and the single machine linear algebra computing system that existing R language platform provides then is limited to its function, is merely able to
It is calculated using single computer.The result of evaluation and test is referring to fig. 2.In Fig. 2, solid line indicates the prototype software that the present invention realizes
The evaluation result of system, dotted line indicate the evaluation result for the linear computing system of single machine that existing R language platform provides.With ginseng
Add the increase of the matrix size of calculating, for the prototype software system that the present invention realizes when completing identical calculating task, institute is time-consuming
Between it is fewer than previously described single machine linear algebra computing system.The prototype software system that evaluation and test shows that the present invention realizes is carrying out
When extensive matrix multiplication operation, calculate time-consuming shorter, calculating speed faster, it was demonstrated that method proposed by the present invention it is effective
Property, demonstrate beneficial effects of the present invention.
Claims (3)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510755923.2A CN105389220B (en) | 2015-11-09 | 2015-11-09 | A Method of Parallel Linear Algebra Computation in Interactive R Language Platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510755923.2A CN105389220B (en) | 2015-11-09 | 2015-11-09 | A Method of Parallel Linear Algebra Computation in Interactive R Language Platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105389220A CN105389220A (en) | 2016-03-09 |
| CN105389220B true CN105389220B (en) | 2019-02-15 |
Family
ID=55421527
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510755923.2A Active CN105389220B (en) | 2015-11-09 | 2015-11-09 | A Method of Parallel Linear Algebra Computation in Interactive R Language Platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105389220B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022483B (en) * | 2016-05-11 | 2019-06-14 | 星环信息科技（上海）有限公司 | The method and apparatus converted between machine learning model |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5845120A (en) * | 1995-09-19 | 1998-12-01 | Sun Microsystems, Inc. | Method and apparatus for linking compiler error messages to relevant information |
| CN1339743A (en) * | 2000-08-23 | 2002-03-13 | 国际商业机器公司 | Method and device for computer software analysis |
-
2015
- 2015-11-09 CN CN201510755923.2A patent/CN105389220B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5845120A (en) * | 1995-09-19 | 1998-12-01 | Sun Microsystems, Inc. | Method and apparatus for linking compiler error messages to relevant information |
| CN1339743A (en) * | 2000-08-23 | 2002-03-13 | 国际商业机器公司 | Method and device for computer software analysis |
Non-Patent Citations (2)
| Title |
|---|
| "P_HPF并行编译系统EXTRINSIC过程调用及对ScaLAPACK并行线性代数算法包的支持";谢军;《http://xueshu.baidu.com/s?wd=paperuri%3A%285826e837b9edcac52efc79230d3f0235%29&filter=sc_long_sign&tn=SE_xueshusource_2kduw22v&sc_vurl=http%3A%2F%2Fwww.doc88.com%2Fp-2045556984539.html&ie=utf-8&sc_us=16634907125052348932》;20140802;第3-5章 |
| "p-HPF支持多范例并行计算的并行编译技术";胡长军，等;《计算机学报》;20010731;第24卷(第7期);全文 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105389220A (en) | 2016-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Blockchain-based digital twin sharing platform for reconfigurable socialized manufacturing resource integration | |
| Adorf et al. | Simple data and workflow management with the signac framework | |
| CN108628605A (en) | Stream data processing method, device, server and medium | |
| WO2016048732A1 (en) | Cloud-based parallel computation using actor modules | |
| CN110717268B (en) | A Portable Component Unit Encapsulation Method Based on FACE Architecture | |
| US9990216B2 (en) | Providing hypercall interface for virtual machines | |
| CN105260177A (en) | SiPESC platform based Python extension module development method | |
| Kinzer et al. | A computational stack for cross-domain acceleration | |
| Shepovalov et al. | FPGA and GPU-based acceleration of ML workloads on Amazon cloud-A case study using gradient boosted decision tree library | |
| Matthew et al. | GillesPy2: a biochemical modeling framework for simulation driven biological discovery | |
| Medeiros et al. | A gpu-accelerated molecular docking workflow with kubernetes and apache airflow | |
| Suram et al. | Engineering design analysis utilizing a cloud platform | |
| CN105389220B (en) | A Method of Parallel Linear Algebra Computation in Interactive R Language Platform | |
| US10606588B2 (en) | Conversion of Boolean conditions | |
| Wozniak et al. | Interlanguage parallel scripting for distributed-memory scientific computing | |
| Zhang et al. | A low-code development framework for cloud-native edge systems | |
| US9361266B2 (en) | System and method for distributed computing | |
| Wang et al. | Transformer: A new paradigm for building data-parallel programming models | |
| Sanderson et al. | Armadillo: An efficient framework for numerical linear algebra | |
| Diez Dolinski et al. | Distributed simulation of P systems by means of map-reduce: first steps with Hadoop and P-Lingua | |
| Chard et al. | PDACS-A Portal for Data Analysis Services for Cosmological Simulations | |
| Makarov et al. | Supercomputer technologies in social sciences: Existing experience and future perspectives | |
| CN105930262A (en) | Application program user interface automated testing method and electronic device | |
| Skjellum et al. | Object‐oriented analysis and design of the Message Passing Interface | |
| Hai Socheat | Automatic and scalable cloud framework for parametric studies using scientific applications |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Gu Rong Inventor after: Wang Zhaokang Inventor after: Huang Yihua Inventor after: Fan Shiqing Inventor before: Huang Yihua Inventor before: Wang Zhaokang Inventor before: Gu Rong Inventor before: Fan Shiqing |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |