Computational resource control device and control method
The computational resource control device optimizes processing circuits in flow-centric computing by creating efficient function chains based on user criteria and system constraints, enhancing resource management and performance.
Patent Information
- Application Number
- JP2023516950
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-04-28
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2041-04-28
AI Technical Summary
Existing methods for controlling computational resources in flow-centric computing do not efficiently optimize processing circuits for desired computational performance.
A computational resource control device and method that includes an input section, equivalent circuit preparation, function chain creation, and optimization units to determine and optimize processing circuits based on user-specified criteria and system constraints, creating efficient function chains for processing execution.
Enables efficient management and operation of computational resources, optimizing processing time, resource consumption, and throughput by creating flexible and optimized function chains.
Smart Images

Figure 0007582457000001 
Figure 0007582457000002 
Figure 0007582457000003
Abstract
Description
[Technical field]
[0001] The present invention relates to a computing resource control device and a control method for use in a computer system or the like. [Background technology]
[0002] Technological innovation is progressing in many fields, such as machine learning, artificial intelligence (AI), and the Internet of Things (IoT), and services are becoming more advanced and added value is being provided by utilizing various information and data. Information processing infrastructure is essential to perform large amounts of calculations in processing this information and data.
[0003] For example, it is difficult to keep up with the rapid increase in data by simply updating the existing information processing infrastructure, and further progress in the future will require the establishment of "post-Moore technology" that goes beyond Moore's Law (Non-Patent Document 1).
[0004] As a post-Moore technology, for example, flow-centric computing is disclosed in Non-Patent Document 2. Flow-centric computing is different from the conventional concept of computing in which processing is performed where data is located, and proposes a new concept of moving data to a location where a computing function is located and processing the data there.
[0005] To realize this flow-centric computing, not only is a broadband communication network necessary for data movement required, but technology to efficiently control computing resources to achieve the desired computing performance is also required. [Prior art documents] [Non-patent literature]
[0006] [Non-Patent Document 1] “NTT Technology Report for Smart World 2020,” Nippon Telegraph and Telephone Corporation, 2020, https: / / www.rd.ntt / _assets / pdf / techreport / NTT_TRFSW_2020_EN_W.pdf [Non-Patent Document 2] R. Takano and T. Kudoh, “Flow-centric computing leveraged by photonic circuit switching for the post-moore era,” Tenth IEEE / ACM International Symposium on Networks-on-Chip(NOCS), Nara, 2016, pp.1-3. https: / / ieeexplore.ieee. org / abstract / document / 7579339 Summary of the Invention [Problem to be solved by the invention]
[0007] In general flow-centric computing (for example, Non-Patent Document 2), a method for controlling, operating, and managing whether or not computational resources can be secured is disclosed.
[0008] However, no technique is disclosed for optimizing the processing circuit that executes the processing and for efficiently controlling, operating, and managing the computing resources. [Means for solving the problem]
[0009] In order to solve the above-mentioned problems, a computational resource control device according to the present invention includes an input unit to which processing content designated by a user is input, an equivalent circuit preparation unit that collects candidates for equivalent circuits, which are processing circuits having a function for executing a part of the processing content, and outputs them as a group of equivalent circuit candidates, and a function chain creation unit that determines a processing execution circuit from the group of equivalent circuit candidates in accordance with a predetermined criterion, determines the connection order of the processing execution circuits, and outputs a function chain that executes the processing content.a function chain optimization unit that optimizes a function chain based on at least one of a computational resource constraint and a performance optimization condition input to the input unit; Equipped with. In addition, the computational resource control device of the present invention includes an input unit into which processing content specified by a user is input, an equivalent circuit preparation unit that collects candidates for equivalent circuits, which are processing circuits having a function for executing part of the processing content, and outputs them as a group of equivalent circuit candidates, a function chain creation unit that determines a processing execution circuit from the group of equivalent circuit candidates in accordance with predetermined criteria, determines the connection order of the processing execution circuits, and outputs a function chain that executes the processing content, and a dynamic chain control unit that determines whether to regenerate the function chain based on system monitor information input to the input unit.
[0010] Further, a computational resource control method according to the present invention includes the steps of: The computer Processing contents specified by the user and computational resource constraints and performance optimization conditions. a step of inputting a function chain that executes a part of the processing content, a step of collecting candidates for a processing circuit having a function for executing a part of the processing content and outputting the candidates as an equivalent circuit candidate group, a step of determining a processing execution circuit from the equivalent circuit candidate group, and a step of outputting a function chain that executes the processing content. and optimizing the function chain based on the computational resource constraints and / or the performance optimization conditions. . In addition, the computational resource control method of the present invention includes the steps of: a computer receiving system monitor information; determining whether to regenerate a function chain based on the system monitor information; receiving processing content specified by a user; collecting candidate processing circuits having a function for executing part of the processing content and outputting them as a group of equivalent circuit candidates; determining a processing execution circuit from the group of equivalent circuit candidates; and outputting a function chain that executes the processing content. Effect of the Invention
[0011] According to the present invention, it is possible to provide a computing resource control device and a control method that can efficiently control, operate, and manage computing resources. [Brief description of the drawings]
[0012] [Figure 1A] FIG. 1A is a block diagram showing a configuration of a computational resource control device according to a first embodiment of the present invention. [Figure 1B] FIG. 1B is a block diagram showing the configuration of a conventional computational resource control device. [Diagram 2] FIG. 2 is a flow chart for explaining the computational resource control method according to the first embodiment of the present invention. [Diagram 3] FIG. 3 is a diagram for explaining an example of a computational resource control method according to the first embodiment of the present invention. [Figure 4] FIG. 4 is a diagram for explaining an example of a computational resource control method according to the first embodiment of the present invention. [Diagram 5] FIG. 5 is a diagram for explaining an example of a computational resource control method according to the first embodiment of the present invention. [Figure 6] FIG. 6 is a diagram for explaining an example of a computational resource control method according to the first embodiment of the present invention. [Figure 7] FIG. 7 is a block diagram showing a configuration of a computational resource control device according to the second embodiment of the present invention. [Figure 8] FIG. 8 is a flow chart for explaining a computational resource control method according to the second embodiment of the present invention. [Figure 9] FIG. 9 is a block diagram showing a configuration of a computational resource control device according to the third embodiment of the present invention. [Figure 10] FIG. 10 is a flow chart for explaining a computational resource control method according to the third embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0013] <First embodiment> A computational resource control device and control method according to a first embodiment of the present invention will be described with reference to FIGS. 1A to 6. FIG.
[0014] <Configuration of computational resource control device> 1A, the computational resource control device 10 according to this embodiment includes an input unit 11, an equivalent circuit preparation unit 12, and a function chain creation unit 13. The computational resource control device 10 is connected to a computer system such as an arithmetic unit, or functions as a part of the computer system.
[0015] The input unit 11 receives processing content designated by the user.
[0016] The equivalent circuit preparation unit 12 collects candidates for processing circuits (hereinafter referred to as "equivalent circuits") that have a function for executing part of the processing content designated by the user, and outputs them as a group of equivalent circuit candidates.
[0017] As candidates for processing circuits, circuits with different parameters that determine the circuit characteristics, such as processing performance (latency and throughput), consumption of computational resources such as circuit area and FPGA resources, and power consumption, are collected.
[0018] When collecting circuits with different characteristics, the equivalent circuit preparation unit 12 can hold available circuit candidates in advance as a database.
[0019] In addition, a basic design of a circuit may be maintained, and by changing parameters that determine the characteristics of the circuit in the basic design, circuits with equivalent input / output information but different characteristics may be created on a case-by-case basis.
[0020] For example, if the source code for the basic design has a parameter that specifies the number of parallel executions of the internal processing of the circuit, a circuit is created by changing this parameter by a specified number within a specified range.When using an FPGA, the source code is prepared in advance, and the parameters of this source code are changed, and circuits with different characteristics are prepared by executing the processes of circuit synthesis, implementation, and bitstream generation.
[0021] The function chain creation unit 13 determines, in accordance with predetermined criteria, the circuits required to execute the processing content specified by the user (hereinafter referred to as "processing execution circuits") from the group of equivalent circuit candidates output by the equivalent circuit preparation unit 12, determines the connection order of the circuits, and outputs a function chain that executes the processing content specified by the user.
[0022] The predetermined criteria for determining the circuit may be, for example, the processing contents specified by the user. Specifically, there are "processing time (latency) priority", "computational resource consumption priority", "input condition priority", "throughput priority", etc. In addition, a function chain may be generated by applying multiple processing contents as AND conditions according to the priority.
[0023] For comparison, FIG. 1B shows the configuration of a conventional computational resource control device 10_2.
[0024] In the conventional computational resource control device 10_2, the function chain creating unit 13_2 creates a function chain based on the processing contents inputted to the input unit 11_2 without considering the constraints (conditions) of the processing.
[0025] On the other hand, since the computational resource control device 10 according to this embodiment is equipped with the equivalent circuit preparation unit 12, the function chain creation unit 13 can determine the processing execution circuit from the group of equivalent circuit candidates output by the equivalent circuit preparation unit 12 in accordance with a predetermined criterion.
[0026] This allows the user to specify processing details such as "priority on processing time (latency)," "priority on computational resource consumption," "priority on input conditions," and "priority on throughput" as predetermined criteria for determining a circuit.
[0027] In addition, when creating a function chain based on the criterion of "prioritizing processing time (latency)", it has the effect of being able to create a function chain that combines circuits with the shortest processing time, especially when there are no constraints on the amount of computational resources consumed. On the other hand, when there are constraints on the amount of computational resources consumed, it has the effect of being able to create a function chain that combines circuits with the shortest processing time under the constraints.
[0028] <Calculation resource control method> The computational resource control method according to this embodiment will be described with reference to FIG.
[0029] First, the process content designated by the user is input (step S1).
[0030] Next, candidates for processing circuits having a function for executing a part of the input processing content are collected and a group of equivalent circuit candidates is output (step S2).
[0031] Next, a processing execution circuit is determined from the group of equivalent circuit candidates according to a predetermined criterion (step S3).
[0032] Finally, a function chain for executing the processing content designated by the user is created and output by determining the connection order of the processing execution circuits (step S4).
[0033] The creation of a function chain in the control method according to the present embodiment will be described in detail below. Figures 3 to 6 show examples of the configuration of a function chain based on the processing contents designated by the user.
[0034] For example, as shown in Figure 3, if "Processing time (latency) priority" is selected, a function chain is created by combining circuits that minimize the processing time (latency) of each circuit constituting the function chain. In the figure, 13_1 is an example of the function chain configuration before the change, and 13_2 is an example of the function chain configuration after the change.
[0035] Here, if there is enough circuit resource, the pipeline with the longest processing time (Function: A) is changed to a circuit with a short processing time. This reduces latency. Also, the output period can be shortened, improving the output amount per unit time (throughput).
[0036] In addition, the next best function chain is created by modifying one of the circuits that make up the function chain with the shortest processing time, and then combining the circuits with the second shortest processing time (the next best).
[0037] Similarly, a function chain is created by combining circuits in ascending order of processing time. If there are no particular constraints on the amount of computational resources consumed, a function chain combining circuits with the smallest processing time is output. On the other hand, if there are constraints on the amount of computational resources consumed, a function chain combining circuits with the smallest processing time under the constraints is selected and output.
[0038] Also, when "prioritize computational resource consumption" is selected, a function chain is created by combining circuits that minimize the computational resource consumption of each circuit constituting the function chain, as shown in Figure 4. In the figure, 14_1 is an example of the function chain configuration before the change, and 14_2 is an example of the function chain configuration after the change.
[0039] In this case, the processing time of the pipeline (Function: B) that has a margin of processing time is allowed to increase as long as the decrease in throughput performance can be suppressed, and the circuit (Function: B) is changed to a circuit that consumes less computational resources. This makes it possible to reduce the computational resource consumption while maintaining the throughput.
[0040] In addition, the runner-up function chain is created by modifying one of the circuits that make up the function chain with the lowest computational resource consumption, and then combining the circuits with the second-lowest computational resource consumption (the runner-up).
[0041] Similarly, function chains are created by combining circuits in ascending order of computational resource consumption. If there are no particular constraints on processing time, a function chain combining circuits that consume the least amount of computational resources is output. On the other hand, if there are constraints on processing time, a function chain combining circuits that consume the least amount of computational resources under the constraints is selected and output.
[0042] In addition, when "Input condition priority" is selected, circuits are combined to satisfy the constraints of the external input speed for the first circuit that constitutes the function chain. When external input packets are input at 100 microsecond intervals, circuits are combined to create a function chain so that the processing time is such that there is no delay in processing in the subsequent circuits, that is, so that no pipelines exceeding 100 microseconds are included.
[0043] Also, if "Throughput Priority" is selected, a function chain is created by combining circuits so as to maximize the output throughput of the function chain.
[0044] In addition, by modifying one of the circuits that make up the function chain with the maximum output throughput and then combining the circuits with the second highest output throughput (the runner-up), a runner-up function chain is created.
[0045] Similarly, function chains are created by combining circuits in descending order of output throughput. In particular, when there are no constraints on computational resource consumption, a function chain combining circuits that maximize the output throughput of the function chain is output. On the other hand, when there are constraints on computational resource consumption, a function chain combining circuits that maximize the output throughput of the function chain under the constraints is selected and output.
[0046] Furthermore, if the throughput does not change and the increase in processing time is within an acceptable range, a function chain is created by combining circuits that consume the least amount of computational resources.
[0047] When pipeline processing is performed in a function chain, even if the processing time of only some of the circuits that make up the function chain is short, if the processing time of the other circuits is long, the processing time will increase within the allowable range, but the change in output throughput can be considered to be within the margin of error. Therefore, a function chain is created by combining circuits that consume the least amount of computational resources. In this case, the time that each circuit that makes up the function chain is idle can be minimized.
[0048] In addition, the creation of a function chain in the case where a branch process that requires synchronization between processing circuits is included will be described with reference to Fig. 5 and Fig. 6. In the figures, 15_1 and 16_1 are configuration examples of the function chain before the processing circuit is changed, and 15_2 and 16_2 are configuration examples of the function chain after the processing circuit is changed. Also, 15_3 and 16_3 show the mode of branch process between the processing circuits.
[0049] In this case, the overall processing time is measured by creating a function chain so that the branching processes are processed in parallel.
[0050] For example, if there is a surplus of circuit resources, the pipeline with the longest processing time (Function: C) is changed to a circuit with a short processing time, as shown in Figure 5. This makes it possible to reduce latency and improve throughput.
[0051] Also, for example, as shown in Fig. 6, an increase in the processing time of a pipeline (Function: B) that has a margin of processing time is permitted within a range that can suppress a decrease in throughput performance, and the circuit (Function: B) is changed to one that consumes less computational resources. This makes it possible to maintain throughput and reduce computational resource consumption.
[0052] In addition, candidates for processing circuits that configure the function chain may be input information from outside. In this case, the equivalent circuit preparation unit 12 prepares candidates for the equivalent circuit so as to include the candidates for the processing circuit input from outside, and inputs them to the function chain creation unit 13 at the subsequent stage.
[0053] <Effects> According to the computational resource control device and control method of this embodiment, when creating a function chain, it is possible to shorten processing time, suppress consumption of computational resources, and achieve highly flexible computational resource control.
[0054] Furthermore, in the process of creating a function chain by combining multiple candidate processing circuits, the function chain can be created based on the processing content of the function chain specified by the user.
[0055] Furthermore, in function chains that include branching processing, it is possible to reduce consumption of computational resources without changing processing performance.
[0056] <Second embodiment> A computational resource control device and control method according to a second embodiment of the present invention will be described with reference to FIGS.
[0057] <Configuration of computational resource control device> 7, the computational resource control device 20 according to this embodiment includes an input unit 11, an equivalent circuit preparation unit 12, a function chain creation unit 13, and a chain optimization unit 21. In the computational resource control device 20, the equivalent circuit preparation unit 12 and the function chain creation unit 13 are similar to those in the first embodiment, but differ from the first embodiment in that the computational resource control device 20 includes a chain optimization unit 21.
[0058] Along with the processing content designated by the user, computational resource constraints and performance optimization conditions are input to the input unit 11. The computational resource constraints and performance optimization conditions may be input by the user, or may be automatically input from a computer system.
[0059] Here, the computational resource constraint is, for example, the state of the entire computer system (usage status, predicted usage, etc.).
[0060] Moreover, the performance optimization condition is, for example, a designated priority, where "priority" indicates the order of priority of the process for which a chain is to be created in the computer system.
[0061] Moreover, the performance optimization condition is, for example, a request for provision of computational resources from processing units other than the processing unit in question, and a response to this request as to whether or not the provision can be made.
[0062] The function chain optimization unit 21 optimizes the function chain based on the processing contents designated by the user as well as on the computational resource constraints and performance optimization conditions.
[0063] <Calculation resource control method> The computational resource control method according to this embodiment will be described with reference to FIG.
[0064] In the computational resource control method according to the present embodiment, the same steps (steps S1 to S4) as those in the computational resource control method according to the first embodiment are executed, and then the function chain is optimized (step S5). Here, the function chain optimization unit 21 optimizes the function chain as described above.
[0065] The optimization of the function chain in the control method according to the present embodiment will be described in detail below.
[0066] For example, when the usage status of computational resources at any point in time is used to optimize a function chain, if the usage status of computational resources is higher than a predetermined standard, the chain optimization unit 21 suppresses the processing performance to a predetermined performance level so as to reduce the consumption of computational resources in the process.
[0067] On the other hand, when the usage of the computational resources at any point in time is lower than a predetermined standard, the chain optimization unit 21 optimizes the processing performance to meet the desired performance without placing any restrictions on the computational resources consumed by the process.
[0068] In addition, when a computational resource usage prediction based on future prediction is used to optimize a function chain, if the computational resource usage prediction is higher than a predetermined standard, the chain optimization unit 21 suppresses the processing performance to a predetermined performance level so as to reduce the consumption of computational resources in the process.
[0069] On the other hand, when the predicted usage of the computational resources is lower than the predetermined standard, the chain optimization unit 21 optimizes the processing performance so as to satisfy the desired performance without placing any restrictions on the computational resources consumed by the processing.
[0070] In this way, the function chain optimization unit 21 can optimize the function chain based on the computational resource constraints.
[0071] In addition, when optimizing a function chain based on a specified priority, when sufficient computational resources are provided for processing to meet the desired performance, the chain optimization unit 21 optimizes the processing performance to meet the desired performance without placing any restrictions on the computational resources consumed by the processing.
[0072] On the other hand, when sufficient computational resources are not provided for processing to satisfy the desired performance, a request is made to a process with a lower priority so that computational resources are provided from the process with a lower priority.
[0073] In this case, when it is possible to provide computational resources and sufficient computational resources are provided for processing to satisfy the desired performance, the processing performance is optimized to satisfy the desired performance without placing any restrictions on the computational resources consumed by the processing.
[0074] In addition, when it is possible to provide computational resources but insufficient computational resources are provided for processing to satisfy the desired performance, the provided computational resources are added and the computational resources available for the processing are set as an upper limit, and the processing performance is optimized so as to approach the desired performance.
[0075] In addition, when computational resources cannot be provided and sufficient computational resources are not provided for processing to meet the desired performance, the processing performance is optimized so as to approach the desired performance, with the computational resources available for the process as an upper limit.
[0076] In this way, the function chain optimization unit 21 can optimize the function chain based on the performance optimization conditions.
[0077] <Effects> According to the computational resource control device and control method of this embodiment, chains can be optimized based on computational resource constraints or performance optimization conditions, enabling more efficient and flexible computational resource control, such as shortening processing time and suppressing consumption of computational resources.
[0078] In this embodiment, an example has been shown in which a chain is optimized based on either a computational resource constraint or a performance optimization condition, but it is of course possible to optimize a chain based on both a computational resource constraint and a performance optimization condition.
[0079] <Third embodiment> A computational resource control device and control method according to a third embodiment of the present invention will be described with reference to FIGS.
[0080] <Configuration of computational resource control device> The computational resource control device 30 according to this embodiment includes an input unit 11, an equivalent circuit preparation unit 12, a function chain creation unit 13, a chain optimization unit 21, and a dynamic chain control unit 31. In the computational resource control device 30, the equivalent circuit preparation unit 12, the function chain creation unit 13, and the chain optimization unit 21 are similar to those in the second embodiment, but differ from the second embodiment in that the computational resource control device 30 includes a dynamic chain control unit 31.
[0081] The input unit 11 receives input of the usage status of the computational resources of the entire system and the progress of processing in response to a user's processing request. Hereinafter, the usage status of the computational resources of the entire system and the progress of processing in response to a user's processing request are referred to as "system monitor information." As in the second embodiment, computational resource constraints and performance optimization conditions are input along with the processing content designated by the user.
[0082] Here, the system monitor information may be input by a user or may be automatically input from a computer system, and the computational resource constraints and performance optimization conditions may be input by a user or may be automatically input from a computer system.
[0083] The dynamic chain control unit 31 judges whether to regenerate the function chain at a predetermined reference value based on the system monitor information. For example, the regeneration of the function chain is started when the predetermined reference value is exceeded. Here, the predetermined reference value is a value indicating the usage status of the computing resources of the entire system or a value indicating the progress status of the processing for the user's processing request.
[0084] <Calculation resource control method> The computational resource control method according to this embodiment will be described with reference to FIG.
[0085] First, system monitor information is input (step S6).
[0086] Next, the system monitor information is judged based on a predetermined reference value (step S7). If the system monitor information exceeds the predetermined reference value, a function chain is created (steps S1 to S5) in the same manner as in the first or second embodiment.
[0087] If the system monitor information is below or equal to a predetermined reference value, the system monitor information is input again (step S6). Here, the system monitor information may be input continuously or at any time.
[0088] The start of function chain regeneration in the control method according to this embodiment will be described in detail below.
[0089] For example, when the input system monitor information exceeds a predetermined reference value, regeneration of the function chain is initiated to change the processing circuit so as to reduce the computational resources of low-priority processes.
[0090] Additionally, a request may be made to the system for additional computing resources.
[0091] In addition, when implementing a processing circuit using an FPGA as a computational resource, computational resources can be added by increasing the number of available FPGAs.
[0092] In this embodiment, an example is shown in which the process of creating a function chain is started when the reference value is exceeded, but the process of creating a function chain may be started when the reference value is exceeded.
[0093] For example, if the usage falls below a predetermined reference value, function chain regeneration is initiated to change the processing circuit so as to provide computing resources to processes whose usage is tight in other function chain creation processes.
[0094] In addition, the number of available FPGAs may be reduced.
[0095] <Effects> According to the computing resource control device and control method of the present embodiment, computing resources can be allocated based on usage, computing resources can be secured when the load is high, and processing performance can be improved when the load is low.
[0096] In this embodiment, an example of application to the second embodiment has been shown, but it may be applied to the first embodiment. In this case, the computational resource control device includes an input unit 11, an equivalent circuit preparation unit 12, a function chain creation unit 13, and a dynamic chain control unit 31, and produces the same effects.
[0097] In an embodiment of the present invention, when the processing content (parent processing) specified by the user is executed by combining multiple processing circuits (child processing), for example, when the entire function chain is composed of a higher-level function chain (parent processing) and a lower-level function chain (child processing), the user may specify the processing content for the child processing.
[0098] Alternatively, the user may specify the process content only for the parent process, in which case the computer system may automatically select from the process content specified for the parent process and specify the process content for the child process.
[0099] <Extended embodiment> Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-mentioned embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention. In addition, the respective embodiments can be combined in any manner within a range that does not contradict each other. [Industrial Applicability]
[0100] The present invention can be applied to computer systems and the like as a computing resource control device and control method. [Explanation of symbols]
[0101] 10 Computational resource control device 11 Input section 12 Equivalent circuit preparation section 13 Function Chain Creation Department
Claims
1. an input section for inputting processing content designated by a user; an equivalent circuit preparation unit that collects candidates for an equivalent circuit, which is a processing circuit having a function of executing a part of the processing content, and outputs the candidates as an equivalent circuit; a function chain creation unit that determines a processing execution circuit from the group of equivalent circuit candidates in accordance with a predetermined criterion, determines a connection order of the processing execution circuits, and outputs a function chain that executes the processing content; a function chain optimization unit that optimizes a function chain based on at least one of a computational resource constraint and a performance optimization condition input to the input unit; A computing resource control device comprising:
2. An input section into which processing content designated by a user is input; an equivalent circuit preparation unit that collects candidates for an equivalent circuit, which is a processing circuit having a function of executing a part of the processing content, and outputs the candidates as an equivalent circuit; a function chain creation unit that determines a processing execution circuit from the group of equivalent circuit candidates in accordance with a predetermined criterion, determines a connection order of the processing execution circuits, and outputs a function chain that executes the processing content; a dynamic chain control unit that determines whether to regenerate a function chain based on system monitor information input to the input unit; A computing resource control device comprising:
3. 3. The computational resource control device according to claim 2, A computational resource control device characterized in that the regeneration of the function chain is determined based on at least one of a value indicating the usage status of computational resources in the entire system and a value indicating the progress of processing in response to a user's processing request.
4. A computational resource control device according to any one of claims 1 to 3, A computing resource control device, characterized in that the processing execution circuit is determined by the processing content designated by a user.
5. A computer comprising: A step of inputting processing contents, computational resource constraints, and performance optimization conditions designated by a user; collecting candidates for processing circuits having a function for executing a part of the processing content, and outputting the candidates as a group of equivalent circuit candidates; determining a processing execution circuit from the group of equivalent circuit candidates; outputting a function chain for executing the processing content; optimizing the function chain based on said computational resource constraints and / or said performance optimization criteria; A computing resource control method for performing the above.
6. A computer comprising: system monitor information is input; determining whether to regenerate a function chain based on the system monitor information; A step of inputting processing content designated by a user; collecting candidates of processing circuits having a function for executing a part of the processing content, and outputting the candidates as a group of equivalent circuit candidates; determining a processing execution circuit from the group of equivalent circuit candidates; outputting a function chain for executing the processing content; A computing resource control method for performing the above.
Citation Information
Patent Citations
Timing synthesis / Optimization device and method and record medium recording program for timing synthesis / optimization
JP2000035957A
Method for designing integrated circuit device
JP2001056827A
Circuit generation support program and circuit generation support method
JP2010026968A
Semiconductor LSI design device and design method
JP2017224128A
