CN102650981A - Synchronous structure between programmable operational level parallel units - Google Patents

Synchronous structure between programmable operational level parallel units Download PDF

Info

Publication number
CN102650981A
CN102650981A CN2011100467443A CN201110046744A CN102650981A CN 102650981 A CN102650981 A CN 102650981A CN 2011100467443 A CN2011100467443 A CN 2011100467443A CN 201110046744 A CN201110046744 A CN 201110046744A CN 102650981 A CN102650981 A CN 102650981A
Authority
CN
China
Prior art keywords
unit
ppe
data
array
register
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
Application number
CN2011100467443A
Other languages
Chinese (zh)
Inventor
杜慧敏
刘镇弢
蒋林
韩俊刚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xi'an Post & Telecommunication College
Xian University of Posts and Telecommunications
Original Assignee
Xi'an Post & Telecommunication College
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xi'an Post & Telecommunication College filed Critical Xi'an Post & Telecommunication College
Priority to CN2011100467443A priority Critical patent/CN102650981A/en
Publication of CN102650981A publication Critical patent/CN102650981A/en
Pending legal-status Critical Current

Links

Landscapes

  • Multi Processors (AREA)
  • Devices For Executing Special Programs (AREA)

Abstract

The invention provides a novel synchronous structure based on an array structure and an array unit, which are adjacently connected with each other. Each unit in an array is programmable through software, and is named as a programmable processing element (PPE). A connection wire between each unit and the adjacently-connected unit only has four directions such as the east, the west, the south and the north, and each unit is fixedly connected with the adjacently-connected unit. Data synchronization is realized among the PPE units through a flag register. Each PPE unit is named as a local PPE and comprises registers in four different directions such as the east, the west, the south and the north; and the registers are named as directional flag registers. When data in one direction is ready, an instruction (high level or low level) is set for the corresponding directional flag register. A flag is recovered after the adjacent unit is read, so that the data synchronization is completed through a mechanism.

Description

Synchronization structure between the programmable operations level Parallel Unit
One, technical field
The present invention relates to a kind of novelly, belong to field of computer architecture based on the array processor structure of adjacent interconnected and the synchronization structure between the array element.
Two, background technology
Because the continuous progress of semiconductor technology; Transistor size integrated in chip is to tens, and traditional single-processor computers architecture faces this several technological barriers: storage wall (Memory Wall), ILP wall (Instruction Level Parallel Wall), power consumption wall (Power Wall) etc.Existing concurrent computer architecture just part has solved the problems referred to above, but can't tackle the problem of the common brick wall (Red Brick Wall) that occurs under the deep sub-micron technique.The object of the invention is exactly to solve the problems referred to above through novel architecture.
Three, summary of the invention
The present invention proposes a kind of novel based on the array structure of adjacent interconnected and the synchronization structure between the array element, as shown in Figure 1.Each unit in this array through software programmable, is called processing elements PPE able to programme (Programmable Processing Element), explains as follows:
(1) annexation: Fig. 1 is a structure by N*N, and except unit all around, the line between each unit and its adjacent unit has only the East, West, South, North four direction, is a kind of fixed connection relation.
(2) PPE unit: the PPE unit shown in accompanying drawing 1, is a kind of programmable unit through instruction programming; It can be homogeneity; Also can be heterogeneous, have control module, data path among present single CPU, also comprise data-carrier store, command memory or comprise both.
(3) data sync between the PPE unit: PPE realizes synchronization of data through flag register between the unit.Each PPE unit is referred to as local PPE, has comprised four different directions registers of East, West, South, North, is referred to as the Directional Sign register.After the DSR of certain direction, indication (high or low level) is set for the respective direction flag register.Local PPE unit reads its four data on the direction according to indication in the direction register, and after local PPE reading of data, the standard of respective direction flag register is set to reset (high or low level), with this mechanism to accomplish data sync.
(4) data transfer between the PPE unit: the PPE unit has comprised four registers group of four corners of the world four direction, is referred to as the data transfer registers group.After local PPE handled, the requirement according to data transfer direction in the instruction write data in the corresponding transmission registers group.
(5) command memory (Instruction Memory): the order register among Fig. 1 is used for the parallel level instruction of store data or is used to deposit the Call instruction that starts each PE unit.
Four, description of drawings:
Fig. 1 array junctions composition of the present invention;
Fig. 2 units synchronization structural representation of the present invention
Five, embodiment
Specifically introduce concrete working method of the present invention below in conjunction with accompanying drawing.
Fig. 2 is PPE inner structure and synchronization structure synoptic diagram.As shown in Figure 2, said synchronization structure is linking to each other with the ALU of processing unit on the four direction through data line independently up and down.Output data line has port register, so that the storage output data supplies local processing unit or adjacent processing unit to use.There is data selector synchronization structure inside, in order to realize the visit of processing unit built-in command to all directions port register.Have zone bit in the port register, adjacent processing unit is realized adjacent PPE unit through the judgement to zone bit data interaction with communicate by letter.The data sync structure of this adjacent cells is not only relevant with zone bit; Also relevant with instruction local and that adjacent PPE unit is inner; Local PPE confirms according to this locality instruction which direction port register is the data after the operation send into; Be only to transmit data, still transmit data to a plurality of directions to a direction.Whether adjacent PPE unit not only will arrive according to the zone bit judgment data, and the data that also will judge arrival are one or a plurality of direction all has data arrives, and what confirm needs according to instruction simultaneously is the data of that direction.The built-in command that has needs basis to come the direction of data and the value that value determines whether to ignore certain direction.

Claims (1)

1. one kind based on the array processor structure of adjacent interconnected and the synchronization structure between the array element; Comprise: 1) based on the array structure of adjacent interconnected; Be a kind of except unit all around; Line between each unit and its adjacent unit has only the long line of nothing of East, West, South, North four direction, has the structure of the relation of fixed connection.2) the data sync structure between the PPE unit, PPE realizes synchronization of data through flag register between the unit.Each PPE unit is referred to as local PPE, has comprised four different directions registers of East, West, South, North, is referred to as the Directional Sign register.After the DSR of certain direction, indication (high or low level) is set for the respective direction flag register.Local PPE unit reads its four data on the direction according to indication in the direction register, and after local PPE reading of data, the standard of respective direction flag register is set to reset (high or low level), with this mechanism to accomplish data sync.
CN2011100467443A 2011-02-25 2011-02-25 Synchronous structure between programmable operational level parallel units Pending CN102650981A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2011100467443A CN102650981A (en) 2011-02-25 2011-02-25 Synchronous structure between programmable operational level parallel units

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2011100467443A CN102650981A (en) 2011-02-25 2011-02-25 Synchronous structure between programmable operational level parallel units

Publications (1)

Publication Number Publication Date
CN102650981A true CN102650981A (en) 2012-08-29

Family

ID=46692989

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2011100467443A Pending CN102650981A (en) 2011-02-25 2011-02-25 Synchronous structure between programmable operational level parallel units

Country Status (1)

Country Link
CN (1) CN102650981A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106502923A (en) * 2016-09-30 2017-03-15 西安邮电大学 In array processor, cluster memory storage accesses ranks two-stage switched circuit

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5535365A (en) * 1993-10-22 1996-07-09 Cray Research, Inc. Method and apparatus for locking shared memory locations in multiprocessing systems
US5630162A (en) * 1990-11-13 1997-05-13 International Business Machines Corporation Array processor dotted communication network based on H-DOTs
CN101454755A (en) * 2005-05-26 2009-06-10 Vns组合有限责任公司 Computer system with increased operating efficiency
CN101546302A (en) * 2009-05-07 2009-09-30 复旦大学 Interconnection structure of multicore processor and hierarchical interconnection design method based on interconnection structure

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5630162A (en) * 1990-11-13 1997-05-13 International Business Machines Corporation Array processor dotted communication network based on H-DOTs
US5535365A (en) * 1993-10-22 1996-07-09 Cray Research, Inc. Method and apparatus for locking shared memory locations in multiprocessing systems
CN101454755A (en) * 2005-05-26 2009-06-10 Vns组合有限责任公司 Computer system with increased operating efficiency
CN101546302A (en) * 2009-05-07 2009-09-30 复旦大学 Interconnection structure of multicore processor and hierarchical interconnection design method based on interconnection structure

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106502923A (en) * 2016-09-30 2017-03-15 西安邮电大学 In array processor, cluster memory storage accesses ranks two-stage switched circuit
CN106502923B (en) * 2016-09-30 2018-08-24 西安邮电大学 Storage accesses ranks two-stage switched circuit in cluster in array processor

Similar Documents

Publication Publication Date Title
CN101320360B (en) Message queuing system for parallel integrated circuit and related operation method
CN109213706A (en) System and method for accessing the storage equipment as memory
CN103150279B (en) Method allowing host and baseboard management controller to share device
CN107113254A (en) Network on adaptive exchange chip
EP3605544B1 (en) Image processor comprising a shift register with reduced wiring complexity
CN103761075B (en) Coarse granularity dynamic reconfigurable data integration and control unit structure
CN202602678U (en) Expansion circuit with multiple CAN (Controller Area Network) channel interfaces
CN102063408A (en) Data bus in multi-kernel processor chip
CN101504632B (en) DMA data transmission method and system, DMA controller
CN101986265A (en) Method for distributing instructions in parallel based on Atom processor
CN101415270A (en) Mobile terminal and communication method for processing module thereof
US7840826B2 (en) Method and apparatus for using port communications to switch processor modes
CN102650981A (en) Synchronous structure between programmable operational level parallel units
CN101720040B (en) Video decoding optimizing method fusing high speed memory and DMA channel
CN103577347A (en) Method for operating memory device, and system for memory operation
CN105718392B (en) Cellular array document storage system and its file-storage device and file memory method
CN103729320A (en) Method for implementing CY7C68013 communication on basis of FPGA (field programmable gate array)
CN102799549A (en) Multi-source-port data processing method and device
EP1588276A1 (en) Processor array
US11604650B1 (en) Packing conditional branch operations
CN202795364U (en) Dynamically reconfigurable test measuring instrument
CN105630400A (en) High-speed massive data storage system
CN209149287U (en) Big data operation acceleration system
CN203706197U (en) Coarse-granularity dynamic and reconfigurable data regularity control unit structure
CN203982379U (en) For the multimode data transmission connectors of coarseness dynamic reconfigurable array

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C53 Correction of patent of invention or patent application
CB02 Change of applicant information

Address after: 710121 Shaanxi city of Xi'an province Changan District Wei Guolu Xi'an University of Posts and Telecommunications

Applicant after: Xi'an University of Posts & Telecommunications

Address before: 710121 Shaanxi city of Xi'an province Changan District Wei Guolu Xi'an University of Posts and Telecommunications

Applicant before: Xi'an Post & Telecommunication College

COR Change of bibliographic data

Free format text: CORRECT: APPLICANT; FROM: XI AN POST + TELECOMMUNICATION COLLEGE TO: XI AN UNIVERSITY OF POSTS + TELECOMMUNICATIONS

C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20120829