Extending the transaction level modeling approach for fast communication architecture exploration

Currently, system-on-chip (SoC) designs are becoming increasingly complex, with more and more components being integrated into a single SoC design. Communication between these components is increasingly dominating critical system paths and frequently becomes the source of performance bottlenecks. It, therefore, becomes imperative for designers to explore the communication space early in the design flow. Traditionally, system designers have used Pin-Accurate Bus Cycle Accurate (PA-BCA) models for early communication space exploration. These models capture all of the bus signals and strictly maintain cycle accuracy, which is useful for reliable performance exploration but results in slow simulation speeds for complex, designs, even when they are modeled using high-level languages. Recently, there have been several efforts to use the Transaction-Level Modeling (TLM) paradigm for improving simulation performance in BCA models. However, these transaction-based BCA (T-BCA) models capture a lot of details that can be eliminated when exploring communication architectures. In this paper, we extend the TLM approach and propose a new transaction-based modeling abstraction level (CCATB) to explore the communication design space. Our abstraction level bridges the gap between the TLM and BCA levels, and yields an average performance speedup of 120% over PA-BCA and 67% over T-BCA models, on average. The CCATB models are not only faster to simulate, but also extremely accurate and take less time to model compared to both T-BCA and PA-BCA models. We describe the mechanisms that produce the speedup in CCATB models and also analyze how the achieved simulation speedup scales with design complexity. To demonstrate the effectiveness of using CCATB for exploration, we present communication space exploration case studies from the broadband communication and multimedia application domains.

In this paper the authors present some key concepts concerning the specification language (PSL) utilization in a system level verification flow for system on chip (SoC) designs. As transaction level modeling (TLM) is the de-facto reference model for SoC design flow, the authors evaluate PSL adoption in TLM context. How to save time and effort in the verification phase during system development steps and how to overcome global system verification limitations through a compositional approach are discussed. Two PSL-based techniques, properties re-use and properties refinement, are described and compared in terms of refinement effort and simulation speed delay

In this work, a system level design and conception of a System-on-a-Chip (SoC) for the execution of cognitive agents in robotics will be presented. The cognitive model of the Concurrent Autonomous Agent (CAA), which was already successfully applied in several robotics applications, is used as a reference for the development of the hardware architecture. This cognitive model comprises three levels that run concurrently, namely the reactive level (perception-action cycle that executes predefined behaviours), the instinctive level (receives goals from cognitive level and uses a knowledge based system for selecting behaviours in the reactive level) and the cognitive level (planning). For the development of such system level hardware model, the C++ library SystemC with Transaction Level Modelling (TLM) 2.0 will be used. A system model of a module that executes a knowledge based system is presented, followed by a system level description of a processor dedicated to the execution of the Graphplan planning algorithm. The buses interconnecting these modules are modelled by the TLM generic payload. Results from simulated experiments with complex knowledge bases for solving planning problems in different robotics contexts demonstrate the correctness of the proposed architecture. Finally, a discussion on performance gains takes place in the end.

Transaction level (TL) models are key to early design exploration, performance estimation and virtual prototyping. Their speed and accuracy enable early and rapid System-on-Chip (SoC) design evaluation and software development. Most devices have only register transfer level (RTL) models that are too complex for SoC simulation. Abstracting these models to TL ones, however, is a challenging task, especially when the RTL description is too obscure or not accessible. This work presents a methodology for automatically creating a TL model of an RTL memory controller component. The device is treated as a black box and a multitude of simulations is used to obtain results, showing its timing behavior. The results are classified into conditional probability distributions, which are reused within a TL model to approximate the RTL timing behavior. The presented method is very fast and highly accurate. The resulting TL model executes approximately 1200 times faster, with a maximum measured average timing offset error of 7.66%.

Transaction-Level Models (TLM) are used for the early validation of embedded software. A TL model is a virtual prototype of the hardware part of a System-on-a-Chip (SoC). When using SystemC for transaction level modeling, the main parallel entities of the hardware platform (processors, DMAs, bus arbiters, etc.) are modeled by asynchronous processes, which are scheduled at simulation time. The specification of this scheduling mechanism is non-deterministic; the set of all possible schedulings of the parallel activities represents the physical parallelism faithfully. Moreover TL models may contain loose timing annotations (intervals for instance), and the set of all possible values of time in these intervals is also meant to represent the hardware behaviors faithfully. However, any simulation engine is built on a deterministic scheduler, and at runtime will use specific values in the time intervals. This means that only a very small subset of all the possible schedulings and timings are exhibited during simulation. Some bugs may be missed if they are due to some behaviors of the hardware that are represented by other schedulings or timings. For a given finite test scenario, the set of valid schedulings and timings of a model is finite, but far too large to be explored fully. We present a solution to cover the set of schedulings and timings efficiently. Our solution is based on dynamic partial order reduction and constraint solving techniques. It gives a complete scheduling and timing set, which guarantees the detection of all local errors and deadlocks for a fixed test scenario.

Intel® CoFluent™ Studio is a visual model-driven development (MDD) solution for creating executable specifications of complex systems. It can be used at any point of the project lifecycle for modeling and validating any electronic or information systems in any application domain: hardware block, software stack, System-on-Chip (SoC), mixed hardware/software embedded system, networked/distributed system, end-to-end Internet-of-Things (IoT) infrastructure and Big Data networks. Intel CoFluent Studio can predict performance data from the application and use cases model execution on a multicore/multiprocessor platform model. Intel CoFluent Studio is a system modeling and simulation toolset based on Eclipse. Models are captured in graphical diagrams using Intel CoFluent optimized domain-specific language (DSL) or standard UML notations—a combination of SysML and the MARTE profile. ANSI C or C++ is used as action language to capture data types and algorithms. Non-functional system requirements or model calibration data such as execution durations, power, or memory values, are added through model attributes. Models are translated into transaction-level modeling (TLM) SystemC code for execution. The SystemC code is instrumented and generates traces that can be monitored with various analysis tools. Fast host-based simulations allow designers to observe the real-time execution of their application models on multiprocessor/multicore platform models. Performance figures such as latencies, throughputs, buffer levels, resource loads, power consumption, memory footprint, and cost can be extracted.

Soc is a technology that integrates heterogeneous system components such as microprocessors, memory logic and DSP's into a single chip. The overall performance of SoC design depends on efficient on-chip communication architectures. Efficient interconnection architecture is necessary interprocessor communication, communication between processors and peripherals and between processor and memory. The communication architecture should be flexible in such a way that it should adapt to various traffic conditions. Currently on-chip interconnection networks are mostly implemented using shared buses which are the most common medium. The arbitration plays a crucial role in determining performance of bus-based system, as it assigns priorities, with which processor is granted the access to the shared communication resources. In the conventional arbitration algorithms there are some drawbacks such as bus starvation problem and low system performance. Hence in this paper, probability based dynamically configurable Round robin arbiters are proposed to handle the discrepancy of existing arbitration algorithms. This bus provides each component a flexible and utmost share of on-chip communication bandwidth and improves the latency in access of the shared bus. The performance of SoC is improved using this probabilistic round robin algorithm with regard to the parameters, latency compared to conventional bus arbitration algorithms.

Data Processing Units for scientific space mission need to process ever higher volumes of data and perform ever complex calculations. But the performance of available space-qualified general purpose processors is just in the lower three digit megahertz range, which is already insufficient for some applications. As an alternative, suitable processing steps can be implemented in hardware on a space-qualified SRAM-based FPGA. However, suitable devices are susceptible against space radiation. At the Institute for Communication and Network Engineering a fault-tolerant, network-based communication architecture was developed, which enables the construction of processing chains on the basis of different processing modules within suitable SRAM-based FPGAs and allows the exchange of single processing modules during runtime, too. The communication architecture and its protocol shall isolate non SEU mitigated or just partial SEU mitigated modules affected by radiation-induced faults to prohibit the propagation of errors within the remaining System-on-Chip. In the context of an ESA study, this communication architecture was extended with further components and implemented in a representative hardware platform. Based on the acquired experiences during the study, this work analyses the actual fault-tolerance characteristics as well as weak points of this initial implementation. At appropriate locations, the communication architecture was extended with mechanisms for fault-detection and fault-differentiation as well as with a hardware-based monitoring solution. Both, the former measures and the extension of the employed hardware-platform with selective fault-injection capabilities for the emulation of radiation-induced faults within critical areas of a non SEU mitigated processing module, are used to evaluate the effects of radiation-induced faults within the communication architecture. By means of the gathered results, further measures to increase fast detection and isolation of faulty nodes are developed, selectively implemented and verified. In particular, the ability of the communication architecture to isolate network nodes without SEU mitigation could be significantly improved.

With power becoming a major constraint for multi-processor embedded systems, it is becoming important for designers to characterize and model processor power dissipation. It is critical for these processor power models to be useable across various modeling abstractions in an electronic system level (ESL) design flow, to guide early design decisions. In this paper, we propose a unified processor power modeling methodology for the creation of power models at multiple granularity levels that can be quickly mapped to an ESL design flow. Our experimental results based on applying the proposed methodology on an OpenRISC processor demonstrate the usefulness of having multiple power models. The generated models range from very high-level two-state and architectural/ISS models that can be used in transaction level models (TLM), to extremely detailed cycle-accurate models that enable early exploration of power optimization techniques. These models offer a designer tremendous flexibility to trade off estimation accuracy with estimation/simulation effort.

With power becoming a major constraint for multiprocessor embedded systems, it is becoming important for designers to characterize and model processor power dissipation. It is critical for these processor power models to be useable across various modeling abstractions in an electronic system level (ESL) design flow, to guide early design decisions. In this paper, we propose a unified processor power modeling methodology for the creation of power models at multiple granularity levels that can be quickly mapped to an ESL design flow. Our experimental results based on applying the proposed methodology on the OpenRISC and MIPS processors demonstrate the usefulness of having multiple power models. The generated models range from very high-level two-state and architectural/instruction set simulator models that can be used in transaction level models, to extremely detailed cycle-accurate models that enable early exploration of power optimization techniques. These models offer a designer tremendous flexibility to trade off estimation accuracy with estimation/simulation effort.

This paper presents Lotterybus, a novel high-performance communication architecture for system-on-chip (SoC) designs. The Lotterybus architecture was designed to address the following limitations of current communication architectures: (i) lack of control over the allocation of communication bandwidth to different system components or data flows (e.g., in static priority based shared buses), leading to starvation of lower priority components in some situations, and (ii) significant latencies resulting from variations in the time-profile of the communication requests (e.g., in time division multiplexed access (TDMA) based architectures), sometimes leading to larger latencies for high-priority communications.We present two variations of Lotterybus: the first is a low overhead architecture with statically configured parameters, while the second variant is a more sophisticated architecture, in which values of the architectural parameters are allowed to vary dynamically.Our experiments investigate the performance of the Lotterybus architecture across a wide range of communication traffic characteristics. In addition, we also analyze its performance in a 4x4 ATM switch sub-system design. The results demonstrate that the Lotterybus architecture is (i) capable of providing the designer with fine grained control over the bandwidth allocated to each SoC component or data flow, and (ii) well suited to provide high priority communication traffic with low latencies (we observed upto 85.4\% reduction in communication latencies over conventional on-chip communication architectures).

The on-chip communication architecture is a primary determinant of overall performance in complex system-on-chip (SoC) designs. Since the communication requirements of SoC components can vary significantly over time, communication architectures that dynamically detect and adapt to such variations can substantially improve system performance. In this paper, we propose Flexbus, a new architecture that can efficiently adapt the logical connectivity of the communication architecture and the components connected to it. Flexbus achieves this by dynamically controlling both the communication architecture topology, as well as the mapping of SoC components to the communication architecture. This is achieved through new dynamic bridge by-pass, and component remapping techniques. In this paper, we introduce these techniques, describe how they can be realized within modern on-chip buses, and discuss policies for run-time reconfiguration of Flexbus-based architectures.The techniques underlying Flexbus are general, and are applicable to a variety of bus standards. We have implemented Flexbus as an extension of the popular AMBA AHB bus, and have evaluated it using a commercial design flow. We report on experiments conducted to analyze its area, timing, and performance under a wide variety of system-level traffic profiles. We have applied Flexbus to two example SoC designs: 1) an IEEE 802.11 MAC processor and 2) an UMTS turbo decoder. Our results show that Flexbus provides gains of up to 34.55 % in application data-rates over conventional architectures, with negligible area overhead and a 3.2% penalty in clock period.

This chapter aims to present a new modeling paradigm that promises to significantly increase the efficiency of developing enterprise information systems. Currently, the software industry faces considerable challenges as it tries to build larger, more complex, software systems with fewer resources. Although modern programming languages such as C++ and Java have in general improved the software development process, they have failed to significantly increase developer’s productivity. Thus, developers are considering other paths to address this issue. One of the potential paths is designing, developing and deploying enterprise information systems using the Model Driven Architecture (MDA). MDA is a model-centric approach that allows for modeling the overall business of an enterprise and capturing requirements to developing, deploying, integrating, and managing different kinds of software components without considering any particular implementation technology. At the center of this approach are models; the software development process is driven by constructing models representing the software under development. Code that expresses the implementation of the model in a certain underlying technology is obtained as a result of model transformation. Thus, the intellectual investment spent in developing the business model of an enterprise is not jeopardized by the continuous changes of the implementation technologies. Currently there are two main approaches trying to implement MDA-based tools. One of the approaches is based on the Object Constraint Language and the other on Action Language. An example of designing, developing and deploying an application using this new modeling paradigm is presented. The MDA approach to software development is considered as the biggest shift since the move from Assembler to the first high level languages.

Neural networks (NNs) are used in numerous applications such as audio–video processing and image classification. NNs can address the limitations of the traditional computer algorithms. Recently, the research utilizing field-programmable gate arrays (FPGAs) to implement NNs is on the rise due to low power dissipation, easy and fast reconfigurability offered by these platforms. This is due to the research and development efforts put into the design optimization to improve the throughput of FPGAs. The integration of appropriate hardware with advanced artificial intelligence (AI) software presents many challenges. This paper offers a generalized model of NNs on a FPGA device. In addition, it presents the design process for the high-level synthesis (HLS) tools to implement the project on FPGA. This paper shows the results of latency, timing, pre- and post-synthesis, implementation, and hardware utilization on Xilinx FPGA target device. In this paper, NN for XOR logic gate operation is performed and verified by using HLS tool, and FPGA implementation is proposed. The NN and backpropagation algorithm are developed in high-level programming languages like Python, Java, C, C++. The training part of the NN has been done by using open-source software Dev C++, and tuned weights are taken to Xilinx Vitis HLS. Simulation, synthesis, and implementation are performed by using Xilinx Vitis HLS and Vivado 2020.1 electronic design automation (EDA) tools. NN architecture with tuned weights is implemented on Xilinx ZYNQ FPGA target device. This paper shows that we were able to achieve latency in one clock cycle interval through pipeline and array partition.KeywordsHigh-level synthesis (HLS)Neural networks (NNs)Register transfer level (RTL)Field-programmable gate arrays (FPGAs)Systems on chip (SOC)Artificial intelligence (AI)Graphic processing units (GPUs)Generalized NNs (GNN)Recurrent neural network (RNN)Feed-forward neural networks (FFNNs)

Chip fabrication technology continues to plunge deeper into sub-micron levels requiring hardware designers to utilize ever-increasing amounts of logic and shorten design time. Toward that end, high-level languages such as C/C++ are becoming popular for hardware description and synthesis in order to more quickly leverage complex algorithms. Similarly, as logic density increases due to technology, power dissipation becomes a progressively more important metric of hardware design. PACT HDL, a C to HDL compiler, merges automated hardware synthesis of high-level algorithms with power and performance optimizations and targets arbitrary hardware architectures, particularly in a System on a Chip (SoC) setting that incorporates reprogrammable and application-specific hardware. PACT HDL is intended for applications well suited to custom hardware implementation such as image and signal processing codes. By making the compiler modular and flexible, optimizations may be executed in any order and at different levels in the compilation process. PACT HDL generates industry standard HDL codes, such as RTL Verilog and VHDL, which may be synthesized and profiled for power using commercial tools. This is the first paper on the PACT compiler project in a series. The compiler framework and introductory optimizations are presented. Later papers will focus on these and other optimizations in detail.