CN108304343A - A kind of chip-on communication method of complexity SOC - Google Patents
A kind of chip-on communication method of complexity SOC Download PDFInfo
- Publication number
- CN108304343A CN108304343A CN201810129340.2A CN201810129340A CN108304343A CN 108304343 A CN108304343 A CN 108304343A CN 201810129340 A CN201810129340 A CN 201810129340A CN 108304343 A CN108304343 A CN 108304343A
- Authority
- CN
- China
- Prior art keywords
- opencl
- communications
- calling
- data stored
- state
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F15/00—Digital computers in general; Data processing equipment in general
- G06F15/16—Combinations of two or more digital computers each having at least an arithmetic unit, a program unit and a register, e.g. for a simultaneous processing of several programs
- G06F15/163—Interprocessor communication
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
- H04L12/42—Loop networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention relates to a kind of chip-on communication methods of complexity SOC, signal to be solved is received including primary processor, the Processing Algorithm of signal to be solved is decomposed into multiple openCL to call, and it is sent in communications and data stored reservoir, the multiple openCL received calling is sent in sequence in communications loop by communications and data stored reservoir again, multiple openCL calling return to communications and data stored reservoir after each signal processing module in communications loop is handled, and are summarized by communications and data stored reservoir and are sent in primary processor.The present invention is based on the scheduling of the unidirectional cyclic structure of communication and data stored reservoir, it can solve telecommunication management when there are a large amount of signal processing modules, the processing module in complicated SoC can be enable to ensure efficiency of transmission when being communicated, the difficulty of star and bus topology structure when chip designs is avoided, and solve multiple processing modules to complete competition when the same openCL is dispatched and collision problem.
Description
Technical field
The present invention relates to communication data treatment technology more particularly to a kind of chip-on communication methods of complexity SOC.
Background technology
The fast development of semiconductor fabrication process so that more massive hardware resource can be integrated on chip, and SoC tends to
In the multi-core heterogeneous system chip by a plurality of types of CPU, DSP and application oriented hardware acceleration engine, and for multiprocessing
The communication of intermodule, existing star and bus topology structural plane can have competition and conflict when handling a large amount of signal processings
Problem, and keep host processor load excessive, the task processing capacity of primary processor is reduced, in addition, existing communication structure
Design difficulty is high, is unfavorable for extending and open, therefore is badly in need of being redesigned for SOC communications so that from based on calculating
Design is changed into the design based on communication, and realizes expansible communication construction.Towards open, the standard that heterogeneous system is general
Change, the communication construction of unified scheduling sum has vast potential for future development.
Invention content
In order to solve the above technical problems, The present invention provides a kind of chip-on communication methods of complexity SOC, including following step
Suddenly：
S1, when primary processor receives signal to be solved, then the Processing Algorithm of signal to be solved is decomposed into multiple openCL
It calls, is sent in communications and data stored reservoir；
After S2, communications and data stored reservoir receive multiple openCL calling, multiple openCL calling is sent to communication successively
On loop；
OpenCL calling is passed uni-directionally on each signal processing module by S3, communications loop；
S4, signal processing module receive openCL and call and handled, and after openCL calling is handled well, re-send to communication
On loop；
S5, the openCL handled well calling are returned to communications loop in communications and data stored reservoir.
Further, the chip-on communication method of the complexity SOC is described based on a kind of on chip communication architecture of complexity SOC
The on chip communication architecture of complicated SOC includes primary processor, communications and data stored reservoir, the communications loop being unidirectionally closed and multiple letters
Number processing module, the communications and data stored reservoir and the equal carry of multiple signal processing modules are in communications loop, the main place
Reason device is connect with communications and data stored reservoir.
Signal processing module is used to run the specific calculating of openCL.
Further, the step S2 further includes：
Store the calling queue that openCL is called.
Further, further include after the step S1：
Communications and data stored reservoir is intercepted in primary processor startup, when listening to the calling queue in communications and data stored reservoir
For sky when, then confirm that current all openCL calling are finished.
Further, openCL calling includes state tag, and the state tag is including being not carried out state and complete
At state.
Further, in step s 2, described multiple openCL calling is sent in communications loop successively is also wrapped before
It includes：
The state tag that openCL calling is arranged is to be not carried out state.
Further, in step s 4, the signal processing module receives openCL and calls and handled, when
After openCL calling is handled well, further include：
It is set as the openCL state tags called that state is completed.
Further, the signal processing module includes idle state and calling state, the signal processing module processing
It is calling state when openCL is called, is otherwise idle state, the communications and data stored reservoir and multiple signal processing modules
For the node in communications loop；When signal processing module is in idle condition, communications loop is monitored, when being not carried out state
When openCL calling is passed through, the information for intercepting openCL calling is handled, and it is to call that current demand signal processing module, which is arranged,
The openCL passed through in communications loop calling will be directly transmitted to down by state when signal processing module is in calling state
One node.
Further, communications and data stored reservoir reads the state tag letter called comprising openCL that communications loop returns
Breath, and will be in this information update to the correspondence openCL calling called in queue.
Further, further include after the step S5：
When the calling queue in communications and data stored reservoir is empty, then confirms that current all openCL calling are finished, lead to
Letter and data management pond summarize that each signal processing module handles as a result, then being communicated with primary processor.
The invention has the advantages that：
1, the communication means of complexity SoC provided by the invention can be realized in chip communicates between multi-processing module, simplifies communication
Complexity, the unidirectional cyclic structure based on communication and data stored reservoir are dispatched, can be solved when there are a large amount of signal processing modules
Telecommunication management, the processing module in complicated SoC can be enable to ensure efficiency of transmission when being communicated, avoid star and
Difficulty of the bus topology structure when chip designs.
2, the communication means of complexity SoC provided by the invention by be based on communications and data stored reservoir by openCL call according to
The secondary method by signal processing module can solve competition and punching when multiple processing modules can complete the same openCL scheduling
Prominent problem, simplifies arbitration mechanism.
Description of the drawings
Fig. 1 is the on chip communication architecture schematic diagram for the complicated SOC that the present invention has 5 signal processing modules；
Fig. 2 is that an openCL is invoked at the flow chart being performed in one-way communication loop；
Fig. 3 is the operational flow diagram on one-way communication loop after the completion of an openCL calling executes.
Specific implementation mode
The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, so that advantages and features of the invention are more
It is easily readily appreciated by one skilled in the art, to make apparent define to protection scope of the present invention.
Embodiment 1：
The chip-on communication method of complexity SOC in the present embodiment a kind of includes the following steps：
S1, when primary processor receives signal to be solved, then the Processing Algorithm of signal to be solved will be decomposed into multiple
OpenCL is called, and is sent in communications and data stored reservoir；
After S2, communications and data stored reservoir receive multiple openCL calling, multiple openCL calling is sent to communication successively
On loop；
OpenCL calling is passed uni-directionally on each signal processing module by S3, communications loop；
S4, signal processing module receive openCL and call and handled, and after openCL calling is handled well, re-send to communication
On loop；
S5, the openCL handled well calling are returned to communications loop in communications and data stored reservoir.
When implementation, a signal processing job analysis is the calling of multiple openCL, all calling quilts by primary processor
The openCL that primary processor sends the storage of communications and data stored reservoir to calls queue, and communications and data stored reservoir is by openCL tune
With being distributed to signal processing module, signal processing module is responsible for running the specific calculating of openCL, and signal processing module can be by
Standardized openCL interfaces call.
The present embodiment main purpose is that providing one kind for the multinuclear isomery processing chip with multiple processing modules is based on
Standardized openCL calls encapsulation, and the chip-on communication method for calling processing module to be handled.Complicated SoC be designed to by
One primary processor, a communication and data stored reservoir, a unidirectional closure communications loop and multiple signal processing models composition.
Unidirectional cyclic structure based on communication and data stored reservoir is dispatched, and communication tube when there are a large amount of signal processing modules can be solved
Reason, can enable the processing module in complicated SoC to ensure efficiency of transmission when being communicated, star and type trunk is avoided to open up
Difficulty of the structure when chip designs is flutterred, openCL scheduling is safeguarded based on communications and data stored reservoir and processing module jointly
State tag method, can solve multiple processing modules can complete competition when the same openCL is dispatched and collision problem.
Embodiment 2：
The chip-on communication method of complexity SOC in the present embodiment a kind of includes the following steps：
SA1, when primary processor receives signal to be solved, then the Processing Algorithm of signal to be solved is decomposed into multiple openCL
It calls, is sent in communications and data stored reservoir；
Communications and data stored reservoir is intercepted in SA2, primary processor startup, when listening to the calling in communications and data stored reservoir
When queue is empty, then confirm that current all openCL calling are finished；
After SA3, communications and data stored reservoir receive multiple openCL calling, the calling queue that storage openCL is called, and set
The state tag for setting openCL calling is to be not carried out state, and then multiple openCL calling is sent in communications loop successively；
OpenCL calling includes state tag, and the state tag includes being not carried out state and state being completed；
OpenCL calling is passed uni-directionally on each signal processing module by SA4, communications loop；
SA5 signal processing modules receive openCL and call and handled, and after openCL calling is handled well, openCL is called
State tag be set as that state is completed, re-send in communications loop；
SA6, the openCL handled well calling are returned to communications loop in communications and data stored reservoir, and communications and data stored reservoir is read
Take communications loop return comprising openCL call state tag information, and by this information update to call queue in correspondence
During openCL is called.
Embodiment 3：
The chip-on communication method of complexity SOC in the present embodiment a kind of includes the following steps：
SB1, when primary processor receives signal to be solved, then the Processing Algorithm of signal to be solved is decomposed into multiple openCL
It calls, is sent in communications and data stored reservoir；
After SB2, communications and data stored reservoir receive multiple openCL calling, the calling queue that storage openCL is called, and set
The state tag for setting openCL calling is to be not carried out state, and then multiple openCL calling is sent in communications loop successively；
OpenCL calling includes state tag, and the state tag includes being not carried out state and state being completed；
OpenCL calling is passed uni-directionally on each signal processing module by SB3, communications loop；
SB4, signal processing module receive openCL and call and handled, after openCL calling is handled well, by openCL tune
State tag is set as that state is completed, and re-sends in communications loop；
SB5, the openCL handled well calling are returned to communications loop in communications and data stored reservoir, and communications and data stored reservoir is read
Take communications loop return comprising openCL call state tag information, and by this information update to call queue in correspondence
During openCL is called；
SB6, when the calling queue in communications and data stored reservoir is empty, then confirm that current all openCL calling have executed
Finish, communications and data stored reservoir summarize the processing of each signal processing module as a result, then being communicated with primary processor.
Embodiment 4：
Based on above-described embodiment, in the present embodiment, the chip-on communication method of above-mentioned complexity SOC is based on a kind of piece of complexity SOC
What upper communication structure was implemented, the on chip communication architecture of complicated SOC includes primary processor, communications and data stored reservoir, unidirectional closure
Communications loop and multiple signal processing modules, communications and data stored reservoir and the equal carry of multiple signal processing modules are in communication loop
On the road, primary processor is connect with communications and data stored reservoir.Above-mentioned signal processing module includes idle state and calling state, institute
It is calling state to state when signal processing module processing openCL is called, and is otherwise idle state, the communications and data stored reservoir
It is the node in communications loop with multiple signal processing modules；When signal processing module is in idle condition, communication loop is monitored
Road, when the openCL for the state that is not carried out calling is passed through, the information for intercepting openCL calling is handled, and current letter is arranged
Number processing module is calling state, when signal processing module is in calling state, will directly will be passed through in communications loop
OpenCL calling is transmitted to next node.
Embodiment 5：
The chip-on communication method of complexity SOC in the present embodiment a kind of is as shown in Figure 1, describe the complexity using the present invention
The complicated SoC schematic diagrames of the chip-on communication method of SoC.Complexity SoC is by a primary processor, five signal processing modules, one
A communications and data stored reservoir and a unidirectional communications loop that is closed are constituted.Primary processor can access and manage communication and number
According to stored reservoir.The unidirectional start node for being closed communications loop is accessed by communications and data stored reservoir, and communications and data stored reservoir is logical
One way link is crossed to be connected with the signal processing module 1 of next node and carry out unidirectional information transmission, each signal processing
Module is all transmitted with next signal processing module opportunity one way link into row information, the last one signal processing module passes through list
The information of its own or the forwarding of other signal processing modules is received and sent to chain road direction communications and data stored reservoir chain.
Embodiment 6：
Fig. 2 is to be invoked at the flow chart being performed in one-way communication loop using an openCL of the complicated SoC of the present invention.For
SoC assigned tasks to be treated are completed, the modules of SoC need to be sequentially completed following work：
First, primary processor will decompose in the task according to the design feature of complicated SoC and obtain 2 openCL calling, remember respectively
For A and B, the two calling divided need to coordinate to execute that assigned tasks could be completed；A and B is pressed into communications and data management
The task call queue in pond simultaneously is noted as being not carried out shape.A and B calling is sequentially transmitted unidirectional logical by communications and data stored reservoir
Believe on loop；The signal processing module 1 being in idle condition intercepts A and calls；Signal processing module 1 be arranged A call label be
It is executed by module 1；Thereafter, be performed flag state call A signal processing module 2 is sent to by unidirectional communications link, locate
The information received is analyzed in the signal processing module 2 of idle state, detects that the label that this is called is to be performed state, that is, turns
To this information to module 3；The information of tape label is by module forwards to communications and data stored reservoir；Communications and data stored reservoir according to
It is state in executing that the A in queue is called in label information setting.After A is sent on one-way communication loop, B also by communication with
Data management pond is sent on one-way communication loop.This time, it such as calls the repeating process of A similar, there is signal processing module 1 first
The information for calling B is received, signal processing module 1 is in the state that A is called that executes at this time, directly forwards B relevant informations to letter
Number processing module 2, after the signal processing module 2 being in idle condition receives information, interception B is called；Signal processing module 2 is set
The label for setting B calling executes for module 2；Thereafter, be performed flag state call B letter is sent to by unidirectional communications link
Number processing module 3, the signal processing module 3 being in idle condition analyze the information received, detect that the label that this is called is
It is performed state, that is, goes to this information to next module；The information of tape label is by module forwards to communications and data stored reservoir；
Communications and data stored reservoir is arranged according to label information calls the B in queue for state in execution.
Embodiment 7：
Fig. 3 is the operation on one-way communication loop after the completion of being executed using the openCL calling of the complicated SoC of the present invention
Flow chart.Signal processing module 1 in execution task complete A call after, A call execute completion status be packaged after by
Module 1 is sent to one-way communication loop, this information is transmitted to processing module 2 by unidirectional communications link, and module 2 detects this calling
Label be execute completion status, go to this information to module 3, the information of tape label is by module forwards to communications and data management
Pond communications and data stored reservoir executes completion status according to label information and A is cleared out of calling queue.In intercepting communication and number
If it is sky to be sent according to the group processor of stored reservoir state and call queue, then it represents that all calling, which have executed, to be terminated, main place
The signal processing tasks for managing device are completed.
Embodiments of the present invention are explained in detail above in conjunction with attached drawing, but the present invention is not limited to above-mentioned implementations
Mode within the knowledge of a person skilled in the art can also be without departing from the purpose of the present invention
Various changes can be made.
Claims (10)
1. a kind of chip-on communication method of complexity SOC, which is characterized in that include the following steps：
S1, when primary processor receives signal to be solved, then the Processing Algorithm of signal to be solved is decomposed into multiple openCL
It calls, is sent in communications and data stored reservoir；
After S2, communications and data stored reservoir receive multiple openCL calling, multiple openCL calling is sent to communication successively
On loop；
OpenCL calling is passed uni-directionally on each signal processing module by S3, communications loop；
S4, signal processing module receive openCL and call and handled, and after openCL calling is handled well, re-send to communication
On loop；
S5, the openCL handled well calling are returned to communications loop in communications and data stored reservoir.
2. the chip-on communication method of complexity SOC according to claim 1, which is characterized in that the on piece of the complexity SOC is logical
Letter method is included primary processor, is led to based on a kind of on chip communication architecture of complexity SOC, the on chip communication architecture of the complexity SOC
Letter and data management pond, the communications loop that is unidirectionally closed and multiple signal processing modules, the communications and data stored reservoir and more
In communications loop, the primary processor is connect a equal carry of signal processing module with communications and data stored reservoir.
3. the chip-on communication method of complexity SOC according to claim 1, which is characterized in that the step S2 further includes：
Store the calling queue that openCL is called.
4. the chip-on communication method of complexity SOC according to claim 3, which is characterized in that also wrapped after the step S1
It includes：
Communications and data stored reservoir is intercepted in primary processor startup, when listening to the calling queue in communications and data stored reservoir
For sky when, then confirm that current all openCL calling are finished.
5. the chip-on communication method of complexity SOC according to claim 3, which is characterized in that the openCL, which is called, includes
State tag, the state tag include being not carried out state and state being completed.
6. the chip-on communication method of complexity SOC according to claim 5, which is characterized in that in step s 2, it is described successively
Further include before multiple openCL calling is sent in communications loop：
The state tag that openCL calling is arranged is to be not carried out state.
7. the chip-on communication method of complexity SOC according to claim 6, which is characterized in that in step s 4, the letter
Number processing module receives openCL and calls and handled, and after openCL, which is called, to be handled well, further includes：
It is set as the openCL state tags called that state is completed.
8. the chip-on communication method of complexity SOC according to claim 5, which is characterized in that the signal processing module packet
Idle state and calling state are included, the signal processing module processing openCL is calling state when calling, and is otherwise idle shape
State, the communications and data stored reservoir and multiple signal processing modules are the node in communications loop；At signal processing module
When idle state, communications loop is monitored, when the openCL for the state that is not carried out calling is passed through, intercepts the letter of openCL calling
Breath is handled, and it is that calling state will be straight when signal processing module is in calling state that current demand signal processing module, which is arranged,
It connects and the openCL passed through in communications loop calling is transmitted to next node.
9. the chip-on communication method of complexity SOC according to claim 3, which is characterized in that communications and data stored reservoir is read
Take communications loop return comprising openCL call state tag information, and by this information update to call queue in correspondence
During openCL is called.
10. the chip-on communication method of complexity SOC according to claim 3, which is characterized in that also wrapped after the step S5
It includes：
When the calling queue in communications and data stored reservoir is empty, then confirms that current all openCL calling are finished, lead to
Letter and data management pond summarize that each signal processing module handles as a result, then being communicated with primary processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810129340.2A CN108304343A (en) | 2018-02-08 | 2018-02-08 | A kind of chip-on communication method of complexity SOC |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810129340.2A CN108304343A (en) | 2018-02-08 | 2018-02-08 | A kind of chip-on communication method of complexity SOC |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108304343A true CN108304343A (en) | 2018-07-20 |
Family
ID=62865077
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810129340.2A Withdrawn CN108304343A (en) | 2018-02-08 | 2018-02-08 | A kind of chip-on communication method of complexity SOC |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108304343A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112506851A (en) * | 2020-12-02 | 2021-03-16 | 广东电网有限责任公司佛山供电局 | SOC chip architecture construction method for solving multi-core access conflict |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030172189A1 (en) * | 2001-07-02 | 2003-09-11 | Globespanvirata Incorporated | Communications system using rings architecture |
| US20030172190A1 (en) * | 2001-07-02 | 2003-09-11 | Globespanvirata Incorporated | Communications system using rings architecture |
| CN1702858A (en) * | 2004-05-28 | 2005-11-30 | 英特尔公司 | Multiprocessor chip with bidirectional ring interconnection |
| CN101789892A (en) * | 2010-03-11 | 2010-07-28 | 浙江大学 | All-node virtual-channel network-on-chip ring topology data exchange method and system |
| CN102135950A (en) * | 2011-03-10 | 2011-07-27 | 中国科学技术大学苏州研究院 | On-chip heterogeneous multi-core system based on star type interconnection structure, and communication method thereof |
| CN102780628A (en) * | 2012-07-31 | 2012-11-14 | 中国人民解放军国防科学技术大学 | On-chip interconnection network routing method oriented to multi-core microprocessor |
| CN102999385A (en) * | 2012-11-06 | 2013-03-27 | 苏州懿源宏达知识产权代理有限公司 | Multiprocessor co-processing method in computation equipment |
| CN103106174A (en) * | 2011-11-09 | 2013-05-15 | 深圳市德赛微电子技术有限公司 | Complex system on-chip (SOC) communication method |
| CN104303174A (en) * | 2012-06-25 | 2015-01-21 | 英特尔公司 | Tunneling platform management messages through inter-processor interconnects |
| CN106209121A (en) * | 2016-07-15 | 2016-12-07 | 中国科学院微电子研究所 | A kind of communications baseband SoC chip of multimode multinuclear |
| CN107171954A (en) * | 2016-03-08 | 2017-09-15 | 华为技术有限公司 | Fault tolerance rout ing method, device and network-on-chip |
-
2018
- 2018-02-08 CN CN201810129340.2A patent/CN108304343A/en not_active Withdrawn
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030172189A1 (en) * | 2001-07-02 | 2003-09-11 | Globespanvirata Incorporated | Communications system using rings architecture |
| US20030172190A1 (en) * | 2001-07-02 | 2003-09-11 | Globespanvirata Incorporated | Communications system using rings architecture |
| CN1702858A (en) * | 2004-05-28 | 2005-11-30 | 英特尔公司 | Multiprocessor chip with bidirectional ring interconnection |
| CN101789892A (en) * | 2010-03-11 | 2010-07-28 | 浙江大学 | All-node virtual-channel network-on-chip ring topology data exchange method and system |
| CN102135950A (en) * | 2011-03-10 | 2011-07-27 | 中国科学技术大学苏州研究院 | On-chip heterogeneous multi-core system based on star type interconnection structure, and communication method thereof |
| CN103106174A (en) * | 2011-11-09 | 2013-05-15 | 深圳市德赛微电子技术有限公司 | Complex system on-chip (SOC) communication method |
| CN104303174A (en) * | 2012-06-25 | 2015-01-21 | 英特尔公司 | Tunneling platform management messages through inter-processor interconnects |
| CN102780628A (en) * | 2012-07-31 | 2012-11-14 | 中国人民解放军国防科学技术大学 | On-chip interconnection network routing method oriented to multi-core microprocessor |
| CN102999385A (en) * | 2012-11-06 | 2013-03-27 | 苏州懿源宏达知识产权代理有限公司 | Multiprocessor co-processing method in computation equipment |
| CN107171954A (en) * | 2016-03-08 | 2017-09-15 | 华为技术有限公司 | Fault tolerance rout ing method, device and network-on-chip |
| CN106209121A (en) * | 2016-07-15 | 2016-12-07 | 中国科学院微电子研究所 | A kind of communications baseband SoC chip of multimode multinuclear |
Non-Patent Citations (2)
| Title |
|---|
| A.J.MARTIN等: "Asynchronous Techniques for System-on-Chip Design", 《PROCEEDINGS OF THE IEEE 》 * |
| 郭桂雨: "基于片上网络多核处理器设计与协同验证", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112506851A (en) * | 2020-12-02 | 2021-03-16 | 广东电网有限责任公司佛山供电局 | SOC chip architecture construction method for solving multi-core access conflict |
| CN112506851B (en) * | 2020-12-02 | 2022-02-11 | 广东电网有限责任公司佛山供电局 | SOC chip architecture construction method for solving multi-core access conflict |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106802826B (en) | Service processing method and device based on thread pool | |
| CN100576177C (en) | Bidirectional grade gridding resource scheduling method based on the QoS constraint | |
| CN108897610A (en) | Method for scheduling task, device, computer equipment and storage medium | |
| CN103019838B (en) | Multi-DSP (Digital Signal Processor) platform based distributed type real-time multiple task operating system | |
| CN106293950A (en) | A kind of resource optimization management method towards group system | |
| CN102662725B (en) | Event-driven high concurrent process virtual machine realization method | |
| CN110377406A (en) | A kind of method for scheduling task, device, storage medium and server node | |
| WO2020232875A1 (en) | Actor model-based task scheduling method and apparatus, and storage medium | |
| CN111858065A (en) | Data processing method, device, storage medium and device | |
| CN109670199A (en) | A kind of efficient power network topology analysis method and device | |
| JPS6122337B2 (en) | ||
| CN108762810B (en) | Network message header processor based on parallel micro-engine | |
| CN111694640B (en) | Data processing method, device, electronic equipment and storage medium | |
| CN109327321B (en) | Network model service execution method and device, SDN controller and readable storage medium | |
| CN108304343A (en) | A kind of chip-on communication method of complexity SOC | |
| CN112395056B (en) | Embedded asymmetric real-time system and electric power secondary equipment | |
| CN114710571A (en) | Data packet processing system | |
| CN107491350B (en) | Interface task call method and device | |
| CN100535864C (en) | Method of invalid time over message under system process scheduling | |
| CN102214094B (en) | Operation is performed via asynchronous programming model | |
| CN101349975B (en) | Method for implementing interrupt bottom semi-section mechanism in embedded operation system | |
| CN114978200B (en) | High-throughput large-bandwidth general channelized GPU algorithm | |
| CN116069480A (en) | Processor and computing device | |
| CN109829005A (en) | A kind of big data processing method and processing device | |
| CN111443898A (en) | Method for designing flow program control software based on priority queue and finite-state machine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180720 |