Design tool for interactive incremental placement of elements on layout plan
通过交互式计算工具和机器学习模型,结合全局和局部放置步骤,解决了芯片设计中NoC拓扑优化的复杂性和效率问题,实现了快速合法放置和设计效率的提升。
Patent Information
- Application Number
- CN202410691267.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-05-29
- Filing Date
- 2024-05-30
- Publication Date
- 2024-12-03
AI Technical Summary
Creating and optimizing network-on-chip (NoC) topology is a complex and time-consuming task in chip design, especially when requirements change, requiring frequent modification of chip layout planning, resulting in the inability to quickly calculate the existing tools during interactive editing. Legal and optimized component placement.
An interactive computing tool was developed, combined with machine learning models, which can edit layout plans in graphical views, satisfy clock domain and power domain constraints through the distinction between locked and unlocked components, and optimize component position using global and local placement steps. , use feedback training models to automate topological analysis and synthesis.
It realizes rapid legal and optimized placement in the topology editing process, reduces design time, improves design efficiency, and continuously improves design quality through the training of machine learning models.
Smart Images

Figure CN119066024A_ABST
Abstract
Description
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS
[0002] This application claims the benefit of U.S. Provisional Application Serial No. 63 / 504,793, filed on May 30, 2023 by Amir CHARIF, entitled “INTERACTIVE INCREMENTAL PLACEMENT ON AFLOORPLAN,” the entire disclosure of which is incorporated herein by reference. Technical Field
[0003] The present technology is in the field of electronic system design and, more particularly, relates to physical implementation guidance for a fast incremental process for computing legal and optimized placement of an entire Network-on-Chip (NoC) topology on a floorplan. Background Art
[0004] Multiprocessor systems have been implemented in System on Chip (SoC) that communicate through NoC. The SoC includes instances of initiator intellectual property (IP) and target IP. Transactions are sent from the initiator to one or more targets in the form of packets using industry standard protocols. An initiator connected to the NoC selects a target using an address, sends a request transaction to the target. The NoC decodes the address and transmits the request from the initiator to the target. The target processes the transaction and sends a response transaction, which is transmitted back to the initiator by the NoC.
[0005] For a given set of performance requirements, such as connectivity and latency between sources and destinations, frequencies of various components, maximum available area for NoC logic and its associated routing (wiring), minimum throughput between sources and destinations, power requirements for the NoC, and the locations of components attached to the NoC on the floorplan, creating an optimal NoC that meets all requirements with the least amount of logic and wiring is a complex task. It is usually the job of a chip architect or chip designer to create the optimal NoC, and it is a difficult and time-consuming task. In addition to being a difficult task, the design of the NoC must be modified every time the requirements change, such as modifying the chip floorplan, adding or removing IP components, or modifying the expected performance. Therefore, this task needs to be repeated frequently during chip design using any tool.
[0006] Consider a software tool for editing a network-on-chip topology on top of a floorplan in a graphical view. A user can load an existing network-on-chip topology that was created without physical information (i.e., without a floorplan), in which case the network elements have no initial positions on the floorplan. The user can also modify the topology (or a portion thereof) on the floorplan, but the component placement is not necessarily legal with respect to multiple floorplan constraints (e.g., clock domain location, power domain location, module location), and the global placement may be non-optimal in terms of wiring length. Therefore, what is needed is a tool that provides a process for interactively computing legal and optimized placement of an existing topology on a floorplan, which process is fast because it is called multiple times during topology editing. Furthermore, there is a need for a tool that utilizes a machine learning model that is trained to perform interactive computations. Summary of the invention
[0007] According to various embodiments and aspects of the present invention, a tool is disclosed that allows interactive computation of legal and optimized placement of existing topology on a floorplan. The tool can be called multiple times during topology editing because it is very fast. The tool also includes a machine learning model that performs the interactive computation. In addition, the tool also has the ability to use feedback to train the machine learning model for automated and assisted topology analysis and synthesis. BRIEF DESCRIPTION OF THE DRAWINGS
[0008] Figure 1 shows a floorplan with a blockage area and a connected zone according to various aspects and embodiments of the present invention;
[0009] Figure 2 shows a floorplan of a blocking region within different power domain constraints and clock domain constraints according to various aspects and embodiments of the present invention;
[0010] Figure 3 A floorplan having network elements requiring global placement according to various aspects and embodiments of the present invention is shown;
[0011] Figure 4 A floorplan with legal placement of components in the floorplan is shown according to various aspects and embodiments of the present invention;
[0012] Figure 5 A floorplan with locally determined legal placements is shown in accordance with various aspects and embodiments of the present invention; and
[0013] Figure 6 A process for placing components in a floorplan according to various aspects and embodiments of the present invention is shown. DETAILED DESCRIPTION
[0014] Various examples of the present technology are described below, which illustrate various aspects and embodiments of the present invention. In general, the examples can use the described aspects in any combination. All statements of the principles, aspects and embodiments and their particular examples herein are intended to cover both their structural equivalents and functional equivalents. Additionally, it is intended that such equivalents include both currently known equivalents and equivalents of future developments, that is, elements of any development that perform the same function, regardless of structure.
[0015] It should be noted that, as used herein, the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise. References in this specification to "an aspect", "aspects", "some aspects", "various aspects", or similar language mean that a particular aspect, feature, structure, or characteristic described in conjunction with any embodiment is included in at least one embodiment of the present invention.
[0016] The phrases "in one embodiment", "in at least one embodiment", "in an embodiment", "in certain embodiments", and similar language that appear throughout this specification may, but do not necessarily, all refer to the same embodiment or similar embodiments. In addition, the aspects and embodiments of the invention described herein are merely exemplary and should not be construed as limiting the scope or spirit of the invention as understood by those of ordinary skill in the art. The disclosed invention is effectively made or used in any embodiment including any novel aspect described herein. All statements describing the principles, aspects, and embodiments of the invention herein are intended to cover both structural equivalents and functional equivalents thereof. Such equivalents are intended to include both currently known equivalents and future developed equivalents.
[0017] As used herein, a transaction can be a request transaction or a response transaction. Examples of request transactions include write requests and read requests.
[0018] As used herein, a node is defined as a distribution point or communication endpoint capable of creating, receiving and / or transmitting information through a communication path or channel. A node may refer to any of the following: a switch, a splitter, a combiner, a buffer, and an adapter. As used herein, splitters and combiners are switches; not all switches are splitters or combiners. As used herein and in accordance with various aspects and embodiments of the present invention, the term "splitter" describes a switch having a single inlet port and multiple outlet ports. As used herein and in accordance with various aspects and embodiments of the present invention, the term "combiner" describes a switch having a single outlet port and multiple inlet ports.
[0019] According to various aspects of the present invention, a system on chip (SoC) includes multiple clock domains and multiple power domains. The clock domain is defined by all logic provided by a given clock input. The clock input is characterized by the clock frequency, which is its most important parameter. The power domain is defined by all logic that obtains power from the same power supply. According to various aspects of the present invention, the power supply is gated, so that a power domain can be turned on or off, or isolated from other power domains. In this way, the designer provides the set of clock domain and power domain constraints as part of the initial design.
[0020] Reference now Figure 1 , showing a floorplan 100 with blocking zones and placement areas (such as connection zone 1 and connection zone 2). According to various aspects of the present invention, there are pinned elements and unlocked elements. The user can use the tool to pin / unpin network elements. A pinned element is an element that will not be moved by the tool's automatic placer or machine learning model; its position serves as a reference point for placing other elements. Unlocked elements can be moved by the tool and can be freely moved in an identified area. According to various aspects of the present invention, the tool includes a placer module. The placer module can change their position for optimization / legalization purposes. According to various aspects of the present invention, network reference points (ports or sockets in fixed positions) can initially be automatically pinned by the tool. A connection zone is a collection of available (i.e., unblocked) floorplan positions such that each position in the collection can reach any other position in the same collection without passing through a blockage or blocking zone. According to various aspects of the present invention, if two network elements are placed in the same connection zone, they are guaranteed to be able to reach each other, that is, the wiring connecting them does not have to cross a blockage.
[0021] Reference now Figure 2, floorplan 100 is shown with different clock domains and power domains, such as clock domain 220 and power domain 210. According to various aspects of the present invention, the tool processes placement constraints for legal placement to meet constraints based on clock requirements and power requirements. As shown in the figure, according to some aspects and embodiments of the present invention, element 204 is located in connection area 1, which includes power domain 210 and clock domain 220. The tool includes a machine learning model that performs a process to determine a legal location for any single network element where it can be placed. For a given location in the floorplan to be legal, several criteria can be checked. According to various aspects of the present invention, one criterion is that the location is in available space and not in a blocked area. According to various aspects of the present invention, one criterion is that if the network element is part of a module with explicit placement restrictions (also called move limits), the location must be within the limits of the restrictions. According to various aspects of the present invention, one criterion is that if the network element is part of a clock domain with explicit placement restrictions, the location must be within the limits of the restrictions. According to various aspects of the present invention, the criteria are applied in the same way to power domain requirements. For example, if element 204 communicates with element 206, and element 206 needs to be placed in connection region 1, then the element must be placed in a location that satisfies both power and clock constraints. In this non-constraining example, location 240 satisfies all constraints and is a legal location to place element 206.
[0022] According to various aspects of the invention, one criterion is associated with locking a network element. If the network element is locked, the only legal position of the unit is its current position; if the current position is not a legal position given the criteria, the tool may report an error. The error is used as feedback to the machine learning model to further train the machine learning model to detect all future errors.
[0023] According to various aspects of the invention, one criterion is that a network element must be able to reach its neighbors, ie the network elements to which it is directly connected; therefore, the network element must be in the same connectivity zone as its neighbors.
[0024] Reference now Figures 3 to 5, the floorplan 100 is changed by a placer module of the tool or a placement module that implements two-stage placement or precisely locates the location in the floorplan 300. All changes and placements are presented in real time through user interaction with the tool using at least one graphical view on a display or graphical user interface. According to various aspects of the present invention, the tool uses a placement process that operates in two stages. According to various aspects of the present invention, the tool uses global coarse-grained placement. The tool places the network on chip (NoC) or network in a global scope; that is, all unlocked cells or elements 310 can be moved simultaneously, away from their current locations, to minimize the global wiring length and the average route length. To compensate for the complexity of global placement, this step does not consider the exact wiring route. According to various aspects of the present invention, the distance between two subsequent network elements is considered to be the Manhattan distance (dX+dY). According to various aspects of the present invention, the tool includes a solver module to solve the problem, which constrains the network elements only to their legal locations and minimizes the total segment length of the sum of the Manhattan distances. According to various aspects of the present invention, the tool utilizes a machine learning model to generate and determine the segment and total segment lengths.
[0025] After this step, according to various aspects and embodiments of the present invention, the tool generates Figure 4 This results in some segments being non-optimal. The total segment (or wiring) length may not be accurate because the exact route is not determined and therefore the total wiring length is not considered. Furthermore, since only the total length is optimized, the machine learning model does not attempt to balance the segment lengths.
[0026] According to various aspects of the invention, the tool edits the initial layout or placement. The tool uses local fine-grained placement to refine and determine the exact result. The tool now operates locally rather than globally to readjust the placement (or reposition) of components along the route that may be non-optimally placed or on a non-optimal path / route. Again, refer to Figure 5, the tool operates locally, placing element 504 and element 506 in floorplan 502 one network element at a time. The tool calculates accurate wiring lengths and takes segment balance into account. According to various aspects of the present invention, the tool utilizes a machine learning model to determine accurate wiring lengths and segment balance. In each iteration, a network element is moved one position (east, west, north, or south) if and only if: the wiring lengths around it are improved (including the total length of the wiring connecting it to its neighboring elements) (considering an accurate straight line path); or the variation in the length of the segments around it is reduced (trying to place the element at an equal distance from all neighbors). Other secondary criteria, such as cell density, can be added here. The tool's machine learning model continues this process until the placement is stable or if the maximum number of iterations is reached. After the tool uses these two complementary steps, the global placement possibility of the network is reached in a short time. After some iterations, the final floorplan of the element is shown in floorplan 510.
[0027] Reference now Figure 6 According to various aspects and embodiments of the present invention, a design tool performs a process 600. At step 602, a floorplan is provided for building a NoC, wherein the floorplan includes a connection area and a blocking area. In addition, the tool receives parameters, requirements and / or constraints of the NoC elements. At step 604, the position of the locked element is identified in each connection area. The tool also receives information about the unlocked element for each connection area and the corresponding locked element to be connected to the unlocked element. At step 606, the tool locates or precisely locates the position of the unlocked element of the corresponding locked element in its corresponding connection area that communicates with it. When the element is placed in the floorplan, the tool determines the legal space for placing each element. These legal spaces are the positions required for placing the unlocked element because these legal spaces meet the requirements or constraints associated with the placement of each unlocked element.
[0028] At step 608, the tool (or the machine learning model of the tool) determines the route connecting the locked element and its corresponding unlocked element. The tool can also identify the routes between other locked elements and other unlocked elements in the same connection area. At step 610, the tool (or the machine learning model) adjusts the position of the unlocked elements in the floorplan to optimize and balance the segments of each route. According to various aspects and embodiments of the present invention, the machine learning model of the tool can also recommend adjustments to the blocking area to create a better floorplan for use.
[0029] According to various aspects of the present invention, the initiator and the target are communicatively connected to the NoC. The initiator is a unit that sends a request. The initiator is usually configured to read and write commands. The target is a unit that serves or responds to requests. The target is usually configured to read and write commands. Each initiator is attached to or attached to the NoC through an NIU. The NIU attached to the initiator is called an initiator network interface unit (INIU). In addition, each target is attached to the NoC through an NIU. The NIU attached to the target is called a target network interface unit (TNIU). The main function of the NoC is to transmit each request from the initiator to the desired destination target, and if the request requires or expects a response, the NoC transmits the response of each target to the corresponding request initiator. The initiator and the target have many different parameters to characterize them. According to various aspects of the present invention, for each initiator and target, the clock domain and power domain to which they belong are defined. The width of the data bus they use to send, write and receive read payloads is several bits. According to various aspects of the present invention, the data bus width of the connection (communication path to / from the target) for sending, writing requests and receiving write responses is also defined. Furthermore, as described herein, the clock domain and power domain definitions are references to the clock domains and power domains present in the previously described SoC.
[0030] According to various aspects of the present invention, it is not required that an initiator be able to send requests to all or multiple targets connected to the NoC. The precise definition of the targets that can receive requests from the initiator is outlined in a connection table not shown in the figure. Connectivity and traffic class labeling information can be represented as an explicit or conceptual matrix. Each initiator has a row and each target has a column. If an initiator must be able to send traffic to a target, there must be a traffic class label at the intersection between the initiator row and the target column. If there is no label at the intersection, the tool does not require a connection between the initiator and the target. According to various aspects of the present invention, the actual format used to represent connectivity can be different, as long as each pair of initiator-target combinations has a precise definition of its traffic class, or there is no classification label if there is no connection. Initiator / target connections that support multiple traffic classes are within the scope of the present invention.
[0031] In the context of this invention, latency refers to the number of clock cycles required for data to pass through a network. Latency can cause problems in a transmission network despite its high bandwidth (frequency). An example of a "real-time traffic class" is video data from a self-driving car camera. Long delays in data propagation are unacceptable. If some data is lost, it makes the vehicle unsafe. A still real-time but less important traffic class is audio and / or video from an entertainment system. Having gaps in the data is undesirable, but not unsafe either. An example of non-real-time data is data from a barometer sensor in a car. It doesn't matter if the data is delayed for a few seconds, since the rate of change of the data is quite slow compared to the operating speed of the SoC.
[0032] According to some aspects of the invention, no scenarios are defined for the tool, in which case the tool optimizes the NoC synthesis process for physical cost, such as lowest gate cost and / or lowest wiring cost.
[0033] According to various aspects of the invention, the tool processes the connectivity matrix with the following defined parameters or components:
[0034] One network interface unit per initiator,
[0035] One network interface unit per target,
[0036] Create a switch for each defined traffic class, called the master switch for that class,
[0037] A switch after each initiator / initiator NIU that separates the traffic to the different master switches that this initiator needs to reach,
[0038] A switch before each target / destination NIU to merge traffic from different master switches sending traffic to that target
[0039] The data width of each switch and the clock domain it belongs to is calculated using the data width of each attached interface and its clock domain as input to the tool. According to various aspects of the invention, each step of converting a network that is part of a NoC also performs calculations of the data width and clock domain of the newly created network elements.
[0040] According to various aspects of the present invention, using the tool, each switch in each merger in the main switch is also decomposed into a cascade of mergers, and each merger in the cascade is placed at a branch point of the merger roadmap of the attached target. The branch point of the roadmap is defined by the fact that the path is divided into two or more branches. The process of decomposing the separator in the cascade of separators retains the original separator function, because the number of inputs of the cascade is still one, and the number of outputs of the cascade is the same as the number of outputs of the original separator. The process of decomposing the merger in the cascade of mergers retains the original merging function, because the number of outputs of the cascade is still one, and the number of inputs of the cascade is the same as the number of inputs of the original merger. According to various aspects of the present invention, the effect of the process is to obtain a set of basic switches, which are represented by mergers and separators, and are physically close to the locations required for the actual connections between the switches.
[0041] According to various aspects of the invention, the tool transforms the network so as to reduce the number of wires used between switches that can be realized while maintaining performance defined in scenarios, which are a set of minimum throughputs required between initiators and targets. According to various aspects of the invention, switches are clustered for performance-aware switching, and mergers and splitters that have been distributed on a roadmap are treated as ordinary switches.
[0042] According to other aspects of the present invention, an extension of the clock domain and power domain on the layout plan is provided, and each component is tested to ensure that it is within the boundaries of the specified clock domain and power domain. If the test fails, the component is moved until a suitable position is found where the test passes. Once a suitable position is found for each component, the routing of each connection between the components is completed. The routing process will find a suitable path for a set of wirings that connect between the components. After the routing is completed, long-distance tube components are inserted on the link as needed based on the time required for the signal to transmit a distance of 1mm, using the provided technical capability information.
[0043] According to some aspects and embodiments of the present invention, the tool generates one or more computer files describing the generated NoC, including:
[0044] List of network elements and their configuration: data width, clock domain.
[0045] The location of each generated network element on the floorplan.
[0046] The set of routes through the network elements that enable connectivity. According to aspects of the invention, a route is an ordered list of network elements, one route per pair (initiator, target), and one route per pair (target, initiator). Routes represent how traffic between pairs flows and through which elements.
[0047] According to various aspects of the invention, the tool is used to generate metrics about the generated NoC, such as: histogram of wire length distribution, number of switches, histogram of switches by size.
[0048] According to another aspect of the invention, the tool automatically inserts various adapters and buffers in the network. The tool inserts adapters based on the required adaptation between two elements with different data widths, different clock domains, and power domains. The tool inserts buffers based on scenarios and detected rate mismatches.
[0049] According to some aspects and embodiments, the tool may be used to ensure that multiple iterations of synthesis are performed for incremental optimization of a NoC, including when one of the constraints provided to the tool is information about a previous run.
[0050] After the synthesis process is performed by the software, the result is obtained in a machine-readable form, such as a computer file that captures the information using a well-defined format. An example of such a format is XML, and another example of such a format is JSON. The scope of the present invention is not limited to a particular format.
[0051] In an embodiment, mechanisms for adapting timing implementation between a first component and a second component within a NoC include, but are not limited to, distance tube insertion, adding buffer stages, switching buffers to higher drive or faster buffers, adjusting wire width, and / or indicating which wire layer(s) are being used.
[0052] Within the scope of the present invention, different mechanisms are physically implemented to overcome timing issues within the synthesized topology, solving timing issues before they occur by passing constraints and / or guidance of the physical implementation. Within the scope of the present invention, the physical implementation of the connection includes, but is not limited to, the physical placement of objects, components, and / or gates around the NoC topology so that downstream tools do not place a first connection far enough away from a second connection to exceed the timing limit. It is more desirable to synthesize a network for all connections that can meet the connectivity requirements and minimize the number of gates in the synthesized network. It is important to minimize the number of gates in the synthesized network, including but not limited to: enabling the NoC to meet the topology parameters; ensuring that timing requirements can be met; building more efficiently in a smaller network; and minimizing the total power used in the interconnect between signals.
[0053] In another embodiment, the components on the driving side are replaced by components with greater driving strength to increase the speed. For example, when a signal is forwarded from a first switch to a second switch, because the distance between the first switch and the second switch is too large, the timing requirement is exceeded, and a distance tube can be placed between the first switch and the second switch, or the driving strength can be increased by using a larger wire with lower resistance to make the signal faster, thereby reducing parasitic elements on a specific line.
[0054] In an embodiment, a floorplanning region is a mechanism that can include, but is not limited to: constraining components such as switches and / or distance tube stages to physical locations and / or approximate physical locations; and / or dividing a logical interconnect into multiple parts that are sized to downstream tools for proper synthesis. The physical configuration and logical configuration of the interconnect are configured together to form a logical interconnect. Floorplanning regions are used to constrain components to specific locations. The need for floorplanning regions is a response to the growing size of the network or interconnect of the SOC being developed. In other words, the network or interconnect of the SOC has grown larger than the capacity of the downstream implementation tools, including but not limited to adding wider interconnects and / or more interconnected components.
[0055] According to various aspects of the invention, based on data associated with signals received and transmitted by the switch, it may be more desirable for the switch to be in a particular restricted location because the performance of the switch may be adversely affected if the switch is moved to a location outside of the restricted location.
[0056] According to various aspects of the present invention, additional constraints can be utilized to place the gates of components such as switches or pipe stages as close together as possible within each individual component. A current problem with existing placement programs is that they tend to place components too far apart from each other. Within the scope of the present invention, a switch is a single component with multiple gates, each of which needs to be placed. While the netlist facilitates the interconnection of these components, downstream tools physically place the components and perform routing to connect the specified logical connections.
[0057] According to various aspects of the present invention, two separate constraints of the first aspect and the second aspect may be used simultaneously. For example, if the components of a switch specified to have a certain performance are distributed over an area larger than the specified area, the performance of the switch will be lower than the originally specified performance. Therefore, it is more desirable to keep the components of the switch close together. In addition, based on a larger network, it is more desirable to keep the components within a specific area or area of the floorplan free space so that the entire network will operate as specified or analyzed.
[0058] In an embodiment, in an approximation in a given semiconductor process, a maximum length of a single wiring is established to ensure that the overall timing of a specific path can be met. In addition, moving from a first component to a second component within a given time is an important aspect. Due to the excessive length between the first component and the second component, it may not be possible to move the signal from the first component to the second component within a given time. Therefore, compared with the prior art, a technical solution of inserting a long-distance tube stage is required to solve this technical problem.
[0059] According to various aspects of the invention, area estimates for first and second components are received and timing and other performance requirements are balanced in an optimization of a NoC to ensure that the resulting NoC gates will be placeable, i.e., they will fit in free space in a floorplan allocated to the NoC.
[0060] According to various aspects of the invention, power requirements for the NoC are also received, and timing and other performance requirements are balanced in the optimization of the NoC to ensure that the resulting NoC does not exceed the power requirements specified in the constraints.
[0061] According to various aspects of the invention, physical routing congestion data is also incorporated, either as an estimate based on the NoC netlist structure or as feedback obtained from downstream implementation tools, and timing and other performance requirements are balanced in the optimization of the NoC to ensure that the resulting NoC will be routable within the free space in the floorplan allocated to the NoC.
[0062] According to some aspects of the present invention, a method for guiding the physical generation of a NoC is disclosed. The tool may include a machine learning model that is trained and receives feedback, the tool receives at least one constraint parameter of the NoC, the at least one constraint parameter is selected from a group of constraint parameters including at least one physical constraint and at least one performance constraint; using the tool, augmenting the physical layout plan of the NoC with information guiding the physical implementation of the NoC; and based on the at least one constraint parameter, using the physical layout plan to constrain the physical implementation of the connection to the location on the physical layout plan.
[0063] Certain methods according to various aspects of the present invention may be performed by instructions stored on a non-transitory computer-readable medium. The non-transitory computer-readable medium stores code including instructions that, if executed by one or more processors, will cause a system or computer to perform the steps of the methods described herein. Non-transitory computer-readable media include: rotating magnetic disks, rotating optical disks, flash random access memory (RAM) chips, and other mechanically movable or solid-state storage media. According to various examples, any type of computer-readable medium is suitable for storing code including instructions.
[0064] Certain examples have been described herein, and it should be noted that different combinations of different components from different examples are possible. The significant features are presented to better explain the examples; however, it is apparent that certain features may be added, modified, and / or omitted without modifying the characteristic aspects of these examples.
[0065] Various examples are methods of behavior using one of the machines or a combination of machines. Method examples are complete anywhere in the world where most of the constituent steps appear. For example, according to various aspects and embodiments of the present invention, IP elements or units include: a processor (e.g., a CPU or GPU), RAM - for example, off-chip dynamic RAM or DRAM, a network interface for wired or wireless connection (such as Ethernet, Wi-Fi, 3G, 4G Long Term Evolution (LTE), 5G and other wireless interface standard radios). The IP may also include various I / O interface devices required for different peripherals (such as touch screen sensors, geolocation receivers, microphones, speakers, Bluetooth peripherals) and USB devices (such as keyboards and mice, etc.). By executing instructions stored in RAM devices, the processor performs the steps of the methods described herein.
[0066] Some examples are one or more non-transitory computer-readable media arranged to store such instructions for the methods described herein. Any machine that stores a non-transitory computer-readable medium including any necessary code can implement the examples. Some examples can be implemented as: a physical device, such as a semiconductor chip; a hardware description language representation of the logical or functional behavior of such a device; and one or more non-transitory computer-readable media arranged to store such hardware description language representations. The description of the principles, aspects, and embodiments described herein covers both structural equivalents and functional equivalents thereof. The elements described herein as coupled have an effective relationship that can be realized by direct connection or indirectly with one or more other intermediate elements.
[0067] Those skilled in the art will recognize many modifications and variations. These modifications and variations include any relevant combination of disclosed features. The description of the principles, aspects, and embodiments described herein covers both their structural equivalents and functional equivalents. The elements described herein as "coupled" or "communicatively coupled" have an effective relationship that can be achieved by direct connection or an indirect connection using one or more other intermediate elements. The embodiments described herein as "communicating with another device, module, or element" include any form of communication or link, and include an effective relationship. For example, a wired connection, wireless protocol, near field protocol, or RFID can be used to establish a communication link.
[0068] To the extent that the terms "including," "includes," "having," "has," "with" or variations thereof are used in the detailed description and claims, such terms are intended to be inclusive in a manner similar to the term "comprising."
[0069] Therefore, the scope of the present invention is not limited to the exemplary embodiments shown and described herein. Rather, the scope and spirit of the present invention are embodied by the appended claims.
Claims
1. A method for guiding physical generation of a network-on-chip (NoC) by iteration of component placement, the method comprising: receiving a floorplan having at least one connection region; receiving a blocked area in the layout plan; positioning at least one locking element in the connection region; identifying a plurality of unlocked elements in communication with the locking element, wherein the plurality of unlocked elements are to be positioned within the connection region; accurately locating a legal placement position for each of the plurality of unlocked elements within the connection region; determining a first routing path for the communication path between the locked element and the plurality of unlocked elements; The placement of the one or more unlocked components within the floorplan is adjusted to improve the first routing path and balance segment lengths until an optimal route is determined. 2 . The method of claim 1 , wherein the legal placement positions for each unlocked component satisfy constraints and requirements for each unlocked component.
3. The method of claim 1, wherein the precise positioning includes initial placement of any unlocked elements relative to the locked elements, thereby creating a segment between locked and unlocked elements.
4. The method of claim 3, wherein at least one of the resulting segments is non-optimal due to placement of unlocked elements.
5. The method according to claim 4, further comprising: The placement of the at least one unlocked element is readjusted to balance the segment.
6. The method of claim 5, wherein the unlocking element is repositioned along an existing segment.
7. The method of claim 1, wherein a machine learning model uses constraints for each of the unlocked components to be placed in the floorplan to identify the legal placement locations for each unlocked component.
8. The method of claim 1, wherein a machine learning model determines segments between each of the unlocked elements and the locked elements based on the placement of the unlocked elements in the floorplan.
9. A design tool for editing a network-on-chip (NoC) topology onto a floorplan in a graphical view, the tool comprising a placement module that identifies global and local placements of components on the floorplan to generate legal placements with optimized routing connections between components, wherein the tool receives the layout plan, blocking regions in the layout plan, connection regions in the layout plan, and positions of locking elements in the layout plan such that at least one connection region includes one or more locking elements, wherein the placement module places a plurality of unlocked components that communicate with corresponding locked components in the floorplan, Wherein the at least one connection region corresponds to a position of a first locked element in communication with a group of unlocked elements selected from the plurality of unlocked elements to generate an initial layout of the unlocked elements.
10. The tool of claim 9, wherein the placement module determines a first routing path for communication between the first locked component and the set of unlocked components.
11. The tool of claim 9, wherein the placement module adjusts placement of the set of unlocked components within at least one connection region of the floorplan to improve a segment of a route from the at least one locked component to the locked component.
12. The tool of claim 9, wherein the placement module comprises a machine learning model trained to place components in a floorplan to build a network-on-chip (NoC).
