阅读说明：本技术 片上网络中的端到端服务质量 (End-to-end quality of service in a network on chip ) 是由 I·A·斯瓦布里克 Y·阿贝尔 M·米陶尔 S·阿玛德 于 2019-01-31 设计创作，主要内容包括：一种生成用于可编程器件中的片上网络(NoC)的配置的示例方法包括：接收(502)多个业务流的业务流需求；基于业务流需求通过NoC为每个业务流分配(508)路由；确定(514)沿着所分配的路由的业务流的仲裁设置；生成(516)用于NoC的编程数据；以及将编程数据加载(518)到可编程器件以配置NoC。(An example method of generating a configuration for a network on chip (NoC) in a programmable device includes: receiving (502) traffic flow requirements for a plurality of traffic flows; assigning (508) a route to each traffic flow through the NoC based on traffic flow demand; determining (514) arbitration settings for traffic flows along the assigned route; generating (516) programming data for the NoC; and loading (518) programming data to the programmable device to configure the NoC.)

1. A method of generating a configuration for a network on chip (NoC) in a programmable device, comprising:

receiving service flow requirements of a plurality of service flows;

allocating a route for each traffic flow through the NoC based on the traffic flow demand;

determining arbitration settings for the traffic flows along the assigned route;

generating programming data for the NoC; and

loading the programming data to the programmable device to configure the NoC.

2. The method of claim 1, wherein receiving the traffic flow demand comprises:

source and destination information for each of the plurality of traffic flows is received.

3. The method of claim 2, wherein receiving the traffic flow demand further comprises:

receiving class information for each of the plurality of traffic flows, wherein the class information includes an assignment of one of a plurality of traffic classes to each of the plurality of traffic flows.

4. The method of claim 3, wherein the step of assigning the route comprises:

selecting a physical channel for each traffic flow of the plurality of traffic flows based on the assigned source and destination; and

selecting a virtual channel for each traffic flow of the plurality of traffic flows based on the assigned traffic class.

5. The method of claim 3, wherein the source and destination information comprises a master circuit and a slave circuit for each traffic flow of the plurality of traffic flows.

6. The method of claim 3, wherein each of the routes is between a master circuit and a slave circuit with one or more switches therebetween.

7. The method of claim 6, wherein each of the one or more switches comprises an arbiter, and wherein determining the arbitration setting comprises assigning a weight to one or more virtual channels input to the arbiter in each of the one or more switches.

8. An integrated circuit, comprising:

a processing system;

a programmable logic region; and

a network on chip (NoC) coupling the processing system and the programmable logic region, the NoC including a master circuit coupled to a slave circuit by one or more physical channels, a first physical channel having a plurality of virtual channels.

9. The integrated circuit of claim 8, wherein each of the plurality of virtual lanes is configured to carry a different class of traffic.

10. The integrated circuit of claim 8, wherein more than one of the plurality of virtual lanes is configured to carry the same class of traffic.

11. The integrated circuit of claims 8 to 10, wherein each of the one or more physical channels comprises a route through one or more switches of the NoC.

12. The integrated circuit of claims 8-11, wherein each of the switches comprises an arbiter having a weight to one or more virtual channels input to the arbiter.

13. The integrated circuit of any of claims 8 to 12, wherein the NoC includes peripheral interconnects configured to program the master circuit, the slave circuit, the physical channels, and the virtual channels.

Technical Field

Examples of the present disclosure relate generally to electronic circuits, and in particular to end-to-end quality of service in a network on chip.

Background

Bus architectures have been found to be unsuitable for certain system-on-a-chip (SoC) integrated circuits (socs). As circuit integration increases, transactions may become blocked and the increase in capacitance creates signaling problems. Instead of a bus structure, a network on chip (NoC) may be used to support data communication between components of the SoC.

Nocs typically include a set of switches that route packets from source circuits ("sources") on the chip to destination circuits ("destinations") on the chip. The layout of the switches in the chip supports the transmission of data packets from a desired source to a desired destination. A data packet may travel through multiple switches in transit from a source to a destination. Each switch may be connected to one or more other switches in the network and route incoming data packets to one of the connected switches or to a destination.

Disclosure of Invention

Techniques for end-to-end quality of service in a network on chip. In one example, a method of generating a configuration for a network on chip (NoC) in a programmable device includes: receiving service flow requirements of a plurality of service flows; distributing a route for each service flow through the NoC based on the service flow requirement; determining arbitration settings for traffic flows along the assigned route; generating programming data for the NoC; and loading programming data to the programmable device to configure the NoC.

In another example, a non-transitory computer readable medium includes instructions stored thereon that are executable by a processor to perform a method of generating a configuration for a network on chip (NoC) in a programmable device, the method comprising: receiving service flow requirements of a plurality of service flows; distributing a route for each service flow through the NoC based on the service flow requirement; determining arbitration settings for traffic flows along the assigned route; generating programming data for the NoC; and loading programming data to the programmable device to configure the NoC.

In another example, an integrated circuit includes: a processing system; a programmable logic region; and a network on chip (NoC) coupling the processing system and the programmable logic region, the NoC including a master circuit coupled to a slave circuit by one or more physical channels, a first physical channel having a plurality of virtual channels.

These and other aspects can be understood with reference to the following detailed description.

Drawings

So that the manner in which the above recited features can be understood in detail, a more particular description, briefly summarized above, may be had by reference to example implementations, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical example implementations and are therefore not to be considered limiting of its scope.

Fig. 1 is a block diagram illustrating a system on a chip (SoC) according to one example.

Fig. 2 is a block diagram illustrating a network on chip (NoC) according to one example.

Fig. 3 is a block diagram illustrating connections between endpoint circuits through a NoC according to one example.

FIG. 4 is a block diagram illustrating a computer system according to one example.

Fig. 5 is a flow diagram illustrating a method of generating configuration data for a NoC, according to one example.

Fig. 6 is a block diagram illustrating a communication system according to one example.

Fig. 7 is a block diagram illustrating arbitration in a switch of a NoC according to an example.

FIG. 8 is a block diagram illustrating assigning weights to virtual channels according to one example.

FIG. 9 is a block diagram illustrating a programmable Integrated Circuit (IC) in which techniques described herein may be employed.

FIG. 10 is a schematic diagram of a Field Programmable Gate Array (FPGA) architecture in which techniques described herein may be employed.

To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements of one example may be beneficially incorporated in other examples.

Detailed Description

Various features are described below with reference to the drawings. It should be noted that the figures may or may not be drawn to scale and that elements of similar structures or functions are represented by like reference numerals throughout the figures. It should be noted that the figures are only intended to facilitate the description of the features. They are not intended as an exhaustive description of the claimed invention or as a limitation on the scope of the claimed invention. Additionally, the illustrated examples need not have all of the aspects or advantages illustrated. Aspects or advantages described in connection with a particular example are not necessarily limited to that example, but may be practiced in any other example, even if not so shown or not explicitly described.

Fig. 1 is a block diagram illustrating a system on a chip (SoC)102 according to one example. The SoC 102 is an Integrated Circuit (IC) that includes a processing system 104, a network on chip (NoC)106, and one or more programmable regions 108. The SoC 102 may be coupled to external circuitry, such as a non-volatile memory (NVM)110 and/or a Random Access Memory (RAM) 112. The NVM 110 may store data that may be loaded into the SoC 102 to configure the SoC 102, such as configuring the NoC106 and the programmable logic region 108. Examples of the processing system 104 and the programmable logic area 108 are described below. Typically, the processing system 104 is connected to the programmable logic region 108 through the NoC 106.

The NoC106 includes end-to-end quality of service (QoS) features for controlling data flows therein. In an example, the NoC106 first separates the data stream into specified traffic classes. Data streams in the same traffic class may share or have independent virtual or physical transmission paths. The QoS scheme applies two priorities between traffic classes. Within and between traffic classes, the NoC106 applies a weighted arbitration scheme to regulate traffic flow and provide bandwidth and latency to meet user requirements. Examples of nocs 106 are discussed further below.

Fig. 2 is a block diagram illustrating a NoC106 according to one example. The NoC106 includes a NoC Master Unit (NMU)202, a NoC Slave Unit (NSU)204, a network 214, a NoC Peripheral Interconnect (NPI)210, and a register (Reg) 212. Each NMU202 is an ingress circuit that connects a master endpoint to the NoC 106. Each NSU 204 is an egress circuit that connects the NoC106 to a slave terminal. The NMU202 is connected to the NSU 204 via a network 214. In one example, the network 214 includes NoC packet switches 206 and routes 208 between the NoC packet switches 206. Each NoC packet switch 206 performs switching of NoC packets. The NoC packet switches 206 are connected to each other and to the NMU202 and the NSU 204 by routes 208 to implement multiple physical channels. In addition, the NoC packet switch 206 supports multiple virtual channels per physical channel. The NPI 210 includes circuitry for programming the NMU202, the NSU 204, and the NoC packet switch 206. For example, the NMU202, the NSU 204, and the NoC packet switch 206 may include registers 212 that determine their functionality. The NPI 210 includes an interconnect coupled to a register 212 to program it to set functionality. Configuration data of the NoC106 may be stored in the NVM 110 and provided to the NPI 210 to program the NoC 106.

Fig. 3 is a block diagram illustrating connections between endpoint circuits through the NoC106, according to one example. In this example, the endpoint circuitry 302 is connected to the endpoint circuitry 304 through the NoC 106. The endpoint circuit 302 is the primary circuit of the NMU202 coupled to the NoC 106. The endpoint circuit 304 is a slave circuit of the NSU 204 coupled to the NoC 106. Each endpoint circuit 302 and 304 may be a circuit in the processing system 104 or a circuit in the programmable logic area 108. Each endpoint circuit in programmable logic region 108 may be a dedicated circuit (e.g., a hardened circuit) or a circuit configured in programmable logic.

The network 214 includes a plurality of physical channels 306. The physical channel 306 is implemented by programming the NoC 106. Each physical channel 306 includes one or more NoC packet switches 206 and associated routes 208. The NMU202 is connected to the NSU 204 by at least one physical channel 306. The physical channel 306 may also have one or more virtual channels 308.

Fig. 4 is a block diagram illustrating a computer system 400 according to one example. Computer system 400 includes a computer 401, input/output (IO) devices 412, and a display 414. Computer 401 includes a hardware platform 418 and software executing on hardware platform 418, including an Operating System (OS)420 and Electronic Design Automation (EDA) software 410. The hardware platform 418 includes a Central Processing Unit (CPU)402, a system memory 408, a storage device ("storage 421"), support circuits 404, and an IO interface 406.

The CPU402 may be any type of general purpose Central Processing Unit (CPU), such as an x 86-based processor, a microprocessor-based processor, a

The processor of (a), etc. The CPU402 may include one or more cores and associated circuitry (e.g., cache memory, Memory Management Unit (MMU), interrupt controller, etc.). CPU402 is configured to execute program code that performs one or more operations described herein and may be stored in system memory 408 and/or storage 421. The support circuits 404 include various devices that cooperate with the CPU402 to manage the flow of data between the CPU402, the system memory 408, the storage 421, the IO interface 406, or any other peripheral devices. For example, the support circuits 404 may include chipsets (e.g., north bridge, south bridge, platform host controller, etc.), voltage regulators, firmware (e.g., BIOS), and so forth. In some examples, the CPU402 may be a System In Package (SiP), system on chip (SoC), or the like, that absorbs all or most of the functionality of the support circuitry 404 (e.g., northbridge, southbridge, or the like).

System memory 408 is a device that allows information, such as executable instructions and data, to be stored and retrieved. System memory 408 may include, for example, one or more Random Access Memory (RAM) modules, such as Double Data Rate (DDR) dynamic RAM (dram). The storage 421 includes local storage (e.g., one or more hard disks, flash memory modules, solid-state disks, and optical disks) and/or a storage interface that enables the computer 401 to communicate with one or more network data storage systems. The IO interface 406 may be coupled to an IO device 412 and a display 414.

OS 420 may be any commodity operating system known in the art, such as

MicrosoftMacAnd the like. A user may interact with the EDA software 410 to generate configuration data for the SoC 102. In particular, the EDA software 410 is configured to generate configuration data for programming the NoC106 to implement various physical and virtual channels for connecting endpoint circuits.

Fig. 5 is a flow diagram illustrating a method 500 of generating configuration data for a NoC106, according to one example. The method 500 may be performed by EDA software 410. The method 500 begins at step 502 where the EDA software 410 receives traffic flow requirements from a user at step 502. In one example, at step 504, EDA software 410 receives source and destination information for each traffic flow (e.g., source and destination endpoints for each traffic flow) specified by a user. A traffic flow is a connection that carries data ("traffic") between endpoints. At step 506, the EDA software 410 receives the class information for each traffic flow specified by the user. Example traffic classes include low latency traffic, isochronous traffic, Best Effort (BE) traffic (e.g., bandwidth guaranteed traffic), and so on.

At step 508, the EDA software 410 assigns a route to each traffic flow through the NoC106 based on the traffic flow requirements. In one example, at step 510, the EDA software 410 selects a physical channel for each traffic flow based on the source and destination of the traffic flow. The NoC106 may have multiple physical routes available between each source and destination. At step 512, EDA software 410 selects a virtual channel for one or more virtual channels based on its traffic class. That is, a given physical channel may have multiple virtual channels and may carry multiple traffic flows separated by traffic classes. Each virtual channel within a physical channel carries only one traffic class, but there are several traffic flows within the same traffic class. For example, a given physical channel may transmit a traffic flow of a low latency traffic class and another traffic flow of an isochronous traffic class in a pair of virtual channels. Note that steps 510 and 512 may occur simultaneously in method 500.

At step 514, EDA software 410 determines the arbitration settings for the traffic stream specified by the user. In one example, EDA software 410 sets the virtual channel with higher priority traffic to have higher priority through switch 206 and sets the virtual channel with lower priority traffic to have lower priority through switch 206. For example, isochronous or low latency traffic may be prioritized over other traffic types. In one example, arbitration uses a deficit scheme. At each arbiter output (e.g., the output of switch 206), there is a combined arbitration for all virtual channels from all input ports to one output port. Each virtual channel of each input port has an independent weight value that provides a specified number of arbitration tokens. Tokens are used to regulate arbitration and control bandwidth allocation between traffic flows. This scheme ensures that all requesters (e.g., endpoints) with tokens are serviced before the tokens are refreshed/reloaded. This ensures that arbitration does not starve because all requests in a group must be serviced before a new group is started. The arbitration settings determined at step 514 may be programmed at boot time or may be dynamically adjusted during operation.

At step 516, the EDA software 410 generates programming data for the NoC 106. The programming data is set to configure the NoC106 to implement physical channels, virtual channels, and optionally arbitration settings. In some examples, the arbitration settings may be dynamically programmed after configuration of the NoC 106. At step 518, the EDA software 410 loads the programming data to the SoC 102 (e.g., by storing the programming data in the NVM 110 or providing data programming directly to the SoC 102).

The method 500 provides a fully programmable end-to-end QoS using the NoC 106. Some socs have relatively fixed interconnects, where flexibility in the arbitration scheme is limited. Other socs have selectable routing and limited QoS priority, but do not have precise bandwidth allocation between individual traffic classes and traffic flows. The method 500 provides a combination of: virtual channels for independent flow control, configurable physical channel routing, deficit arbitration among groups, and assignment of traffic classes.

Fig. 6 is a block diagram illustrating a communication system 600 according to one example. The communication system 600 includes a slave device 604 coupled to a NoC106₀And 604₁Master device 602 (slave device 604)₀……602₄(master device 602). The master device 602 and the slave device 604 comprise endpoint circuitry in the SoC 102 that is coupled to the NMU202 and the NSU 204, respectively. The NoC106 includes a NoC Packet Switch (NPS)206 (e.g., NPS 206)_0,0……206_0,3And NPS206_1,0……206_1,3)。

Master device 602₀And a master device 602₁Coupled to the NPS206_0,0. Master device 602₀Coupling to NPS206 through Low Latency (LL) virtual channels_0,0. Master device 602₁Coupling to NPS206 through Best Effort (BE) virtual channels_0,0. Master device 602₃Coupling to NPS206 through BE virtual channels_0,1. Master device 602₃Coupling to NPS206 through Isochronous (ISOC) virtual channels_0,3. Master device 602₄Coupling to NPS206 through ISOC virtual channels_0,3。NPS 206_0,1Coupled to the NPS206_0,2。NPS206_0,2Coupled to the NPS206_0,3。

NPS 206_0,0Coupled to the NPS206_1,0。NPS 206_0,1Coupled to the NPS206_1,1。NPS 206_1,2And NPS206_1,3Unconnected, and unused in the current configuration of communication system 600. NPS206_1,0Coupled to a slave device 604₀。NPS206_1,1Coupled to a slave device 602₁。NPS 206_1,0Coupled to the NPS206_1,1。

In operation, the master device 602₀To the slave device 604₀Low latency traffic is sent. Master device 602₁And 602₂Are all to slave device 604₀Best effort traffic is sent. Master device 602₃And 602₄To the slave device 604₁Isochronous traffic is transmitted. Each traffic flow enters each switch on a separate physical channel. NPS206_0,0And NPS206_1,0Between, NPS206_0,1And NPS206_1,1And NPS206_1,0And slave device 604₀There are two virtual channels in between (designated by a pair of lines). Other paths use only a single virtual channel on a physical channel (e.g., at the NPS 206)_0,1And NPS206_0,2And NPS206_1,1And slave device 602₁In between). Each NPS206 has output port arbitration that controls the mixing of traffic from the input ports to the output ports, as described further below.

Fig. 7 is a block diagram illustrating arbitration in the switches 206 of the NoC106, according to an example. Each switch 206 includes an arbiter 702. In this example, arbiter 702 includes three input ports designated as input port 0, input port 1, and input port 2. However, switch 206 and arbiter 702 may include any number of input ports. The arbiter 702 includes an output port designated "out".

As shown in fig. 7, in this example, there is no incoming traffic flow at input port 2. Input port 0 has two virtual channels that receive two traffic streams (e.g., one low latency traffic stream and one isochronous traffic stream). Input port 1 has a single virtual channel carrying one traffic flow (e.g., best effort traffic). Each input port of the arbiter 702 has an assigned weight. The weights control the relative share of arbitration bandwidth allocated to each traffic stream. In this example, port 0 has arbitration weights of 4 and 8 for the respective virtual channels, and port 1 has arbitration weight of 4 on a single virtual channel. This means that, in the available bandwidth of the output port, the first traffic flow at port 0 acquires 25% of the bandwidth, the second traffic flow at port 0 acquires 50% of the bandwidth, and the traffic flow at port 1 acquires 25% of the bandwidth. For example, low latency traffic at port 0 (due to higher priority) may be allocated more bandwidth than best effort traffic (lower priority). This means that if all requesters are transmitting, then arbiter 702 will service the low latency traffic as long as it has an arbitration token. Best-effort traffic will get service if it has tokens and no other higher priority requesters also have tokens. If there are requestors and no arbitration tokens remain, the arbitration tokens will be reloaded according to the assigned weights. If all requesters' tokens are exhausted, the arbiter 702 also reloads the arbitration token.

The above description is for one arbitration point. The programming of each arbitration point on a given physical path ensures that there is sufficient end-to-end bandwidth. Using high priority assignments for certain virtual channels ensures that the transaction receives lower latency/lower jitter services. The use of arbitration weights and deficit arbitration ensures that all requestors receive a certain amount of bandwidth according to their arbitration weights for a period of time corresponding to the sum of all arbitration weights. Such a group may have less service time if some requesters do not send traffic.

FIG. 8 is a block diagram illustrating assigning weights to virtual channels according to one example. This example includes two arbiters 702₁And 702₂. Arbiter 702₁Arbitration occurs between the physical channels 802, 804, and 806. Arbiter 702₂Arbitration occurs between physical channels 806, 808, and 810. Each physical channel 802, 804, 806, and 808 includes two virtual channels designated vc0 and vc 1. In this example, there are six different sources (e.g., masters) designated as src0 … … src 5. The source src0 is on vc0 of the physical channel 808. The source src1 is on vc1 of the physical channel 808. The source src2 is on vc0 of the physical channel 802. The source src3 is on vc1 of the physical channel 802. The source src4 is on vc0 of the physical channel 804. The source src5 is on vc1 of the physical channel 804. Arbiter 702₂Programmed to provide weights 10 on vc0 of physical channel 808 and weights 20 on vc1 of physical channel 808. Arbiter 702₂Programmed to provide weights 30 on vc0 of physical channel 806 and weights 40 in vc1 of physical channel 806. Arbiter 702₁Programmed to provide a weight of 10 on vc0 of physical channel 802 and a weight of 30 on vc1 of physical channel 802. Arbiter 702₁Programmed to provide weights 20 on vc0 of the physical channel 804 and weights 10 on vc1 of the physical channel 804. This weighting scheme results in arbiter 702₂At the output, src0 has weight 10, src1 has weight 20, src2 has weight 10, src3 has weight 30, src4 has weight 20, src5 has weight 10. The bandwidth acquired by each source is proportional to its weight. Those skilled in the art will appreciate that various other weighting schemes may be employed between any number of arbiters for any number of sources in a similar manner.

Fig. 9 is a block diagram illustrating a programmable IC 1 according to one example that may be used as an implementation of the SoC 102 shown in fig. 1. Programmable IC 1 includes programmable logic 3, configuration logic 25 and configuration memory 26. The programmable IC 1 may be coupled to external circuits such as a nonvolatile memory 27, a DRAM 28, and other circuits 29. Programmable logic 3 includes logic cells 30, support circuits 31, and programmable interconnects 32. The logic unit 30 includes circuitry that may be configured to implement a general logic function for multiple inputs. The support circuits 31 include specialized circuits such as transceivers, input/output blocks, digital signal processors, memories, and the like. The logic cells and support circuits 31 may be interconnected using programmable interconnects 32. Information for programming the logic cells 30, for setting parameters of the support circuits 31, and for programming the programmable interconnects 32 is stored in the configuration memory 26 by the configuration logic 25. Configuration logic 25 may retrieve configuration data from non-volatile memory 27 or any other source (e.g., DRAM 28 or from other circuitry 29). In some examples, programmable IC 1 includes a processing system 2. The processing system 2 may include a microprocessor, memory, support circuits, IO circuits, and the like.

Fig. 10 shows a Field Programmable Gate Array (FPGA) implementation of the programmable IC 1, which includes a number of different programmable blocks, including a transceiver 37, configurable logic blocks ("CLBs") 33, random access memory blocks ("BRAMs") 34, input/output blocks ("IOBs") 36, configuration and clock logic ("CONFIG/CLOCKS") 42, digital signal processing blocks ("DSPs") 35, dedicated input/output blocks ("I/O") 41 (e.g., configuration ports and clock ports), and other programmable logic 39, such as digital clock managers, analog-to-digital converters, system monitoring logic, and so forth. The FPGA may also include a PCIe interface 40, an analog-to-digital converter (ADC)38, and the like.

In some FPGAs, each programmable block may include at least one programmable interconnect element ("INT") 43, the INT 43 having connections to input and output terminals 48 of programmable logic elements within the same block, as shown in the example included at the top of fig. 10. Each programmable interconnect element 43 may also include connections to interconnect segments 49 of adjacent programmable interconnect elements in the same or other blocks. Each programmable interconnect element 43 may also include connections to interconnect segments 50 of a common routing resource between logic blocks (not shown). The common routing resources may include routing channels between a logic block (not shown) including a track of interconnect segments (e.g., interconnect segment 50) and a switch block (not shown) for connecting the interconnect segments. An interconnect segment (e.g., interconnect segment 50) of the generic routing resource may span one or more logical blocks. The programmable interconnect elements 43 together with general routing resources implement a programmable interconnect structure ("programmable interconnect") for the FPGA shown.

In an example implementation, the CLB 33 may include a configurable logic element ("CLE") 44, which may be programmed to implement user logic, and a single programmable interconnect element ("INT") 43. BRAM 34 may include a BRAM logic element ("BRL") 45, and one or more programmable interconnect elements. Typically, the number of interconnect elements included in a block depends on the height of the block. In the illustrated example, a BRAM block has the same height as five CLBs, but other numbers (e.g., four) may also be used. In addition to an appropriate number of programmable interconnect elements, DSP block 35 may include DSP logic elements ("DSPL") 46. In addition to one instance of programmable interconnect element 43, IOB 36 may include two instances of an input/output logic element ("IOL") 47, for example. It will be clear to those skilled in the art that the actual I/O pads connected to, for example, the I/O logic element 47 are generally not limited to the area of the input/output logic element 47.

In the illustrated example, a horizontal region near the center of the die (as shown in fig. 10) is used to configure, provide clocks and other control logic. Vertical columns 51 extending from this horizontal area or column are used to distribute the clock and configuration signals across the width of the FPGA.

Some FPGAs utilizing the architecture shown in fig. 10 include other logic blocks that disrupt the regular columnar structure making up a large part of the FPGA. The additional logic blocks may be programmable blocks and/or dedicated logic.

Note that fig. 10 is intended only to illustrate an exemplary FPGA architecture. For example, the number of logic blocks in a row, the relative widths of the rows, the number and order of rows, the types of logic blocks included in the rows, the relative sizes of the logic blocks, and the interconnect/logic implementations included at the top of FIG. 10 are purely exemplary. For example, in an actual FPGA, wherever a CLB appears, more than one adjacent CLB row is typically included to facilitate efficient implementation of user logic, but the number of adjacent CLB rows varies with the overall size of the FPGA.

In one example, a method of generating a configuration for a network on chip (NoC) in a programmable device may be provided. Such a method may include: receiving service flow requirements of a plurality of service flows; distributing a route for each service flow through the NoC based on the service flow requirement; determining arbitration settings for traffic flows along the assigned route; generating programming data for the NoC; and loading programming data to the programmable device to configure the NoC.

In such a method, the step of receiving traffic flow requirements may comprise: source and destination information for each of a plurality of traffic flows is received.

In such a method, the step of receiving a traffic flow requirement may further comprise: category information for each of a plurality of traffic flows is received, wherein the category information includes an assignment of one of the plurality of traffic categories to each of the plurality of traffic flows.

In such a method, the step of assigning a route may comprise: selecting a physical channel for each traffic flow of a plurality of traffic flows based on the assigned source and destination; and selecting a virtual channel for each traffic flow of the plurality of traffic flows based on the assigned traffic class.

In such a method, the source and destination information may include a master circuit and a slave circuit for each of a plurality of traffic flows.

In such an approach, each of the routes may be between a master circuit and a slave circuit with one or more switches therebetween.

In such a method, each of the one or more switches may include an arbiter, and wherein the step of determining the arbitration setting comprises assigning a weight to one or more virtual channels input to the arbiter in each of the one or more switches.

In another example, a non-transitory computer-readable medium may be provided having stored thereon instructions executable by a processor to perform a method of generating a configuration for a network on chip (NoC) in a programmable device. Such non-transitory computer readable media has stored thereon instructions executable by a processor to perform a method of generating a configuration for a network on chip (NoC) in a programmable device, the method may include: receiving service flow requirements of a plurality of service flows; distributing a route for each service flow through the NoC based on the service flow requirement; determining arbitration settings for traffic flows along the assigned route; generating programming data for the NoC; and loading programming data to the programmable device to configure the NoC.

In such a non-transitory computer readable medium, the step of receiving traffic flow requirements may comprise: source and destination information for each of a plurality of traffic flows is received.

In such a non-transitory computer readable medium, the step of receiving traffic flow requirements further comprises: category information for each of a plurality of traffic flows is received, wherein the category information includes an assignment of one of the plurality of traffic categories to each of the plurality of traffic flows.

In such a non-transitory computer readable medium, the step of assigning the route may include: selecting a physical channel for each traffic flow of a plurality of traffic flows based on the assigned source and destination; and selecting a virtual channel for each traffic flow of the plurality of traffic flows based on the assigned traffic class.

In such a non-transitory computer readable medium, the source information and the destination information include a master circuit and a slave circuit for each of a plurality of traffic flows.

In such a non-transitory computer readable medium, each of the routes may be between a master circuit and a slave circuit with one or more switches between the master circuit and the slave circuit.

In such a non-transitory computer readable medium, each of the one or more switches may include an arbiter, and wherein the step of determining the arbitration setting may include assigning a weight to one or more virtual channels input to the arbiter in each of the one or more switches.

In another example, an integrated circuit may be provided. Such integrated circuits may include: a processing system; a programmable logic region; and a network on chip (NoC) coupling the processing system and the programmable logic region, the NoC including a master circuit coupled to a slave circuit by one or more physical channels, a first physical channel having a plurality of virtual channels.

In such an integrated circuit, each of a plurality of virtual channels may be configured to carry a different class of traffic.

In such an integrated circuit, more than one of the plurality of virtual channels may be configured to carry the same class of traffic.

In such an integrated circuit, each of the one or more physical channels includes a route through one or more switches of the NoC.

In such an integrated circuit, each of the switches includes an arbiter having a weight for one or more virtual channels input to the arbiter.

In such integrated circuits, a NoC may include peripheral interconnects configured to program a master circuit, a slave circuit, a physical channel, and a virtual channel.

While the foregoing is directed to particular examples, other and further examples may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.