The energy consumption of digital services has become a concern for stakeholders committed to sustainability. Raising awareness of this consumption is essential to improve the energy efficiency of digital services. However, expressing the energy usage of digital services in an easily understandable and actionable way remains a challenge. We address this challenge by proposing a first operational energy labeling method for digital services in the computing continuum. Our approach enables stakeholders, including cloud and network providers, application developers, researchers, and end-users of digital services, to better understand and improve the energy efficiency of their applications. Focusing on video surveillance digital services, and using the enhanced iFogSim framework, we propose an energy labeling scheme, and demonstrate its merits with extensive scenario analysis and simulation. We further discuss how our approach can help reduce energy consumption and/or improve performance, all without modifying the application's functional parameters or system architecture.

Designing next generation distributed CyberPhysical Systems (dCPS) requires effective Design Space Exploration (DSE) methods to evaluate system design alternatives and their impact on performance. While existing DSE approaches focus on hardware optimization and software-to-hardware mapping, they often overlook parallel execution opportunities within software tasks. Current application workload models for complex dCPS assume fixed execution orders, limiting the ability to explore and exploit software parallelism. To address this issue, we propose refined workload models derived from execution traces that capture both inter- and intra-process dependencies. Building on these models, we present a method to identify tasks that can be safely reordered or executed in parallel without modifying the existing software implementation. We validate our approach through a case study on the ASML Twinscan lithography machine, demonstrating measurable performance improvements without impacting the system functional correctness.

Features of microservice architectures, such as scalability, separation of concerns, and their ability to facilitate the rapid evolution of polyglot systems, have made them popular with large organizations employing many software developers. However, the features that make them attractive also create complexity and require maintenance over the evolution of an application, especially concerning application decomposition. Mi-croservice architecture decomposition evolves together with the application and is prone to errors referred to as architectural anti-patterns over its lifetime. These can be difficult to detect and manage because of their informal natural language definitions and a lack of automated tooling. This paper addresses this challenge by proposing an automated methodology for detection of architectural anti-patterns related to microservice dependencies. As a part of this methodology, a novel Granular Hardware Utilization-Based Service Dependency Graph (GHUBS) model is automatically inferred from telemetry data. Three commonly known anti-patterns have been formalized and algorithms provided to detect them in the model. The methodology is supported by an open-source tool that automatically detects and visualizes anti-patterns. The proposed methodology and tool are validated using both synthetic data and a case study of a popular microservice benchmarking suite, showing that instances of the formalized anti-patterns can be successfully detected.

Digital services enable users to interact with a broad range of applications and, as such, have become an essential part of our daily lives. Although convenient, their ubiquity comes at a significant cost in energy, raising sustainability concerns. We access these services by triggering a computing continuum, spanning from the device to the edge, fog, and cloud. Scheduling decisions made at each layer impact the overall quality of service (QoS) and energy consumption of digital services.

Industrial Cyber-Physical Systems (CPS) are com-plex heterogeneous and distributed computing systems, typically integrating and interconnecting a large number of subsystems and containing a substantial number of hardware and software components. Producers of these distributed Cyber-Physical Systems (dCPS) face serious challenges with respect to designing the next generations of these machines and require proper support in making (early) design decisions to avoid expensive and time consuming oversights. This calls for efficient and scalable system-level Design Space Exploration (DSE) methods for dCPS. In this position paper, we review the current state of the art in DSE, and argue that efficient and scalable DSE technology for dCPS is more or less non-existing and constitutes a largely unchartered research area. Moreover, we identify several re-search challenges that need to be addressed and discuss possible directions for targeting such DSE technolozy for dCPS.

Abstract This paper presents results and observations from a survey of 120 industry practitioners in the field of real-time embedded systems. The survey provides insights into the characteristics of the systems being developed today and identifies important trends for the future. It extends the results from the survey data to the broader population that it is representative of, and discusses significant differences between application domains. The survey aims to inform both academics and practitioners, helping to avoid divergence between industry practice and academic research. The value of this research is highlighted by a study showing that the aggregate findings of the survey are not common knowledge in the real-time systems community.

System-Performance Engineering (SysPE) encompasses modeling formalisms, methods, techniques, and industrial practices to design systems for performance, where performance is taken integrally into account during the whole system life cycle. Industrial SysPE state of practice is generally model-based. Due to the rapidly increasing complexity of systems, there is a need to develop and establish model-driven methods and techniques. To structure the field of SysPE, we identify (1) industrial challenges motivating the importance of SysPE, (2) scientific challenges that need to be addressed to establish model-driven SysPE, (3) important focus areas for SysPE and (4) best practices. We conducted a survey to collect feedback on our views. The responses were used to update and validate the identified challenges, focus areas, and best practices. The final result is presented in this paper. Interesting observations are that industry sees a need for better design-space exploration support, more than for additional performance modeling and analysis techniques. Also tools and integral methods for SysPE need attention. From the identified focus areas, scheduling and supervisory control is seen as lacking established best practices.

Automotive software implements different functionalities as multiple control applications sharing common platform resources. Although such applications are often developed independently, the control performance of the resulting system depends on how these applications are integrated. A key integration challenge is to efficiently schedule these applications on shared resources with minimal control performance degradation. We formulate this problem as that of scheduling multiple distributed periodic control tasks that communicate via messages with non-zero jitter. The optimization criterion used is a piecewise linear representation of the control performance degradation as a function of the end-to-end latency of the application. The three main contributions of this article are: 1) a constraint programming (CP) formulation to solve this integration problem optimally on time-triggered architectures; 2) an efficient heuristic called Flexi; and 3) an experimental evaluation of the scalability and efficiency of the proposed approaches. In contrast to the CP formulation, which for many real-life problems might have unacceptably long running times, Flexi returns nearly optimal results (0.5 percent loss in control performance compared to optimal) for most problems with more acceptable running times.

Satisfying real-time requirements in cyber-physical systems is challenging as timing behaviour depends on the application software, the embedded hardware, as well as the execution environment. This challenge is exacerbated as real-world, industrial systems often use unpredictable hardware and software libraries or operating systems with timing hazards and proprietary device drivers. All these issues limit or entirely prevent the application of established real-time analysis techniques.

This paper presents results and observations from a survey of 120 industry practitioners in the field of real-time embedded systems. The survey provides insights into the characteristics of the systems being developed today and identifies important trends for the future. The survey aims to inform both academics and practitioners, helping to avoid divergence between industry practice and fundamental academic research.

Timing analysis of safety-critical real-time embedded systems should be free of both optimistic and pessimistic aspects. The multiframe model was devised to eliminate the pessimism in the schedulability analysis of systems with tasks whose worst-case execution times vary from job to job, according to known patterns. However, this model is optimistic and unsafe for multicores with shared memory controllers, since it ignores memory contention, and existing approaches to stall analysis based on memory regulation are very pessimistic if straightforwardly applied. This paper remedies this by adapting existing stall analyses for memory-regulated systems of conventional Liu-and-Layland tasks to the multiframe model. Experimental evaluations with synthetic task sets (and different task and memory budget assignment heuristics) show up to 85% higher scheduling success ratio for our analysis, compared to the frame-agnostic analysis, enabling higher platform utilisation without compromising safety. We also explore implementation aspects, such as how to speed up the analysis and how to trade off accuracy with tractability.

In mixed-criticality systems, tasks of different criticality share system resources, mainly to reduce cost. Cost is further reduced by using adaptive mode-based scheduling arrangements, such as Vestal’s model, to improve resource efficiency, while guaranteeing schedulability of critical functionality. To simplify safety certification, servers are often used to provide temporal isolation between tasks. In its simplest form, a server is a periodically recurring time window, in which some tasks are scheduled. A server’s computational requirements may greatly vary in different modes, although state-of-the-art techniques and schedulability tests do not allow different budgets to be used by a server in different modes. This results in a single conservative execution budget for all modes, increasing system cost. The goal of this paper is to reduce the cost of mixed-criticality systems through three main contributions: (i) a scheduling arrangement for uniprocessor systems employing fixed-priority scheduling within periodic servers, whose budgets are dynamically adjusted at run-time in the event of a mode change, (ii) a new schedulability analysis for such systems, and (iii) heuristic algorithms for assigning budgets to servers in different modes and ordering the execution of the servers. Experiments with synthetic task sets demonstrate considerable improvements (up to 52.8%) in scheduling success ratio when using dynamic server budgets vs. static “one-size-fits-all-modes” budgets.

Mixed-criticality (MC) multicore system design must reconcile safety guarantees and high performance. The interference among cores on shared resources in such systems leads to unpredictable temporal behaviour. Memory bandwidth regulation among different cores can be a useful tool to mitigate the interference when accessing main memory. However, for mixed-criticality systems conforming to the (well-established) Vestal model, the existing schedulability analyses are oblivious to memory stalling effects, including stalls from memory bandwidth regulation. This makes it unsafe. In this paper, we address this issue by formulating a schedulability analysis for mixed-criticality fixed-priority-scheduled multicore systems using per-core memory access regulation. We also propose multiple heuristics for memory bandwidth allocation and task-to-core assignment. We implement our analysis and heuristics in a tool and evaluate them, performance-wise, through extensive experiments. Our experiments show that stall-oblivious schedulability analysis may be optimistic due to contention on shared memory resources.

In multicore architectures, there is potential for contention between cores when accessing shared resources, such as system memory. Such contention scenarios are challenging to accurately analyse, from a worst-case timing perspective. One way of making memory contention in multicores more amenable to timing analysis is the use of memory regulation mechanisms. It restricts the number of accesses performed by any given core over time by using periodically replenished per-core budgets. Typically, this assumes that all cores access memory via a single shared memory controller. However, ever-increasing bandwidth requirements have brought about architectures with multiple memory controllers. These control accesses to different memory regions and are potentially shared among all cores. While this presents an opportunity to satisfy bandwidth requirements, existing analysis designed for a single memory controller are no longer safe. This work formulates a worst-case memory stall analysis for a memory-regulated multicore with two memory controllers. This stall analysis can be integrated into the schedulability analysis of systems under fixed-priority partitioned scheduling. Five heuristics for assigning tasks and memory budgets to cores in a stall-cognisant manner are also proposed. We experimentally quantify the cost in terms of extra stall for letting all cores benefit from the memory space offered by both controllers, and also evaluate the five heuristics for different system characteristics.

Contains fulltext : 191723.pdf (Publisher’s version ) (Open Access)

Mixed-criticality multicore system design must often guarantee both safety and high performance. Memory bandwidth regulation among different cores can be a useful tool for guaranteeing safety, as it mitigates the interference when accessing main memory. The use of mode changes and system models like Vestal's can help provide both safety, for critical functions, and scheduling performance, by efficiently utilising the platform. This work therefore combines per-core memory access regulation with the well-established Vestal model and improves on the state-of-the-art in two respects: 1) We allow the memory access budgets of the cores to be dynamically adjusted, when the system undergoes a mode change, reflecting the different needs in each mode, for better schedulability. 2) We devise memory-regulation-aware and stall-aware schedulability analysis for such systems, based on AMC-max. By comparison, the state-of-the-art offered no option of dynamic adjustment of core budgets, and only offered regulation-aware schedulability analysis based on AMC-rtb, which is inherently more pessimistic. Finally, 3) we consider different task assignment and bandwidth allocation heuristics, to assess the improvement from the dynamic memory budgets and new analysis. Our results show improvements in schedulability ratio of up to 9.1% over the state-of-the-art.

The complexity of embedded application design is increasing with growing user demands. In particular, automotive embedded systems are highly complex in nature, and their functionality is realized by a set of periodic tasks. These tasks may have hard real-time requirements and communicate over an interconnect. The problem is to efficiently co-schedule task execution on cores and message transmission on the interconnect so that timing constraints are satisfied. Contemporary works typically deal with zero-jitter scheduling, which results in lower resource utilization, but has lower memory requirements. This article focuses on jitter-constrained scheduling that puts constraints on the tasks jitter, increasing schedulability over zero-jitter scheduling. The contributions of this article are: 1) Integer Linear Programming and Satisfiability Modulo Theory model exploiting problem-specific information to reduce the formulations complexity to schedule small applications. 2) A heuristic approach, employing three levels of scheduling scaling to real-world use-cases with 10,000 tasks and messages. 3) An experimental evaluation of the proposed approaches on a case-study and on synthetic data sets showing the efficiency of both zero-jitter and jitter-constrained scheduling. It shows that up to 28 percent higher resource utilization can be achieved by having up to 10 times longer computation time with relaxed jitter requirements.

The design of mixed-criticality systems often involves painful tradeoffs between safety guarantees and performance. However, the use of more detailed architectural models in the design and analysis of scheduling arrangements for mixed-criticality systems can provide greater confidence in the analysis, but also opportunities for better performance. Motivated by this view, we propose an extension of Vestal's model for mixed-criticality multicore systems that (i) accounts for the per-task partitioning of the last-level cache and (ii) supports the dynamic reassignment, for better schedulability, of cache portions initially reserved for lower-criticality tasks to the higher-criticality tasks, when the system switches to high-criticality mode. To this model, we apply partitioned EDF scheduling with Ekberg and Yi's deadline-scaling technique. Our schedulability analysis and scalefactor calculation is cognisant of the cache resources assigned to each task, by using WCET estimates that take into account these resources. It is hence able to leverage the dynamic reconfiguration of the cache partitioning, at mode change, for better performance, in terms of provable schedulability. We also propose heuristics for partitioning the cache in low- and high-criticality mode, that promote schedulability. Our experiments with synthetic task sets, indicate tangible improvements in schedulability compared to a baseline cache-aware arrangement where there is no redistribution of cache resources from low- to high-criticality tasks in the event of a mode change.

Many-core processors can provide the computational power required by future complex embedded systems. However, their adoption is not trivial, since several sources of interference on COTS many-core platforms have adverse effects on the resulting performance. One main source of performance degradation is the contention on the Network-on-Chip (NoC), which is used for communication among the compute cores via the off-chip memory. Available analysis techniques for the traversal time of messages on the NoC do not consider many of the architectural features found on COTS platforms. In this work, we target a state-of-the-art many-core processor, the Kalray MPPAR®. A novel partitioning strategy for reducing the contention on the NoC is proposed. Further, we present an analysis technique dedicated to the proposed partitioning strategy, which considers all architectural features of the COTS NoC. Additionally, it is shown how to configure the parameters for flow-regulation on the NoC, such that the Worst-Case Traversal Time (WCTT) is minimal and buffers never overflow. The benefits of our approach are evaluated based on extensive experiments that show that contention is significantly reduced compared to the unconstrained case, while the proposed analysis outperforms a state-of-the-art analysis for the same platform. An industrial case study shows the tightness of the proposed analysis.

Future real-time embedded systems will increasingly incorporate mixed application models with timing constraints running on the same multi-core platform. These application models are dataflow applications with timing constraints and traditional real-time applications modelled as independent arbitrary-deadline tasks. These systems require guarantees that all running applications execute satisfying their timing constraints. Also, to be cost-efficient in terms of design, they require efficient mapping strategies that maximize the use of system resources to reduce the overall cost.

Embedded systems are increasingly based on multi-core platforms to accommodate a growing number of applications, some of which have real-time requirements. Resources, such as off-chip DRAM, are typically shared between the applications using memory interconnects with different arbitration polices to cater to diverse bandwidth and latency requirements. However, traditional centralized interconnects are not scalable as the number of clients increase. Similarly, current distributed interconnects either cannot satisfy the diverse requirements or have decoupled arbitration stages, resulting in larger area, power and worst-case latency. The four main contributions of this article are: 1) a Globally Arbitrated Memory Tree (GAMT) with a distributed architecture that scales well with the number of cores, 2) an RTL-level implementation that can be configured with five arbitration policies (three distinct and two as special cases), 3) the concept of mixed arbitration policies that allows the policy to be selected individually per core, and 4) a worst-case analysis for a mixed arbitration policy that combines TDM and FBSP arbitration.We compare the performance of GAMT with centralized implementations and show that it can run up to four times faster and have over 51 and 37 percent reduction in area and power consumption, respectively, for a given bandwidth.

There exist many dataflow applications with timing constraints that require real-time guarantees on safe execution without violating their deadlines. Extraction of timing parameters (offsets, deadlines, periods) from these applications enables the use of real-time scheduling and analysis techniques, and provides guarantees on satisfying timing constraints. However, existing extraction techniques require the transformation of the dataflow application from highly expressive dataflow computational models, for example, Synchronous Dataflow (SDF) and Cyclo-Static Dataflow (CSDF) to Homogeneous Synchronous Dataflow (HSDF). This transformation can lead to an exponential increase in the size of the application graph that significantly increases the runtime of the analysis. In this article, we address this problem by proposing an offline heuristic algorithm called slack-based merging . The algorithm is a novel graph reduction technique that helps in speeding up the process of timing parameter extraction and finding a feasible real-time schedule, thereby reducing the overall design time of the real-time system. It uses two main concepts: (a) the difference between the worst-case execution time of the SDF graph’s firings and its timing constraints (slack) to merge firings together and generate a reduced-size HSDF graph, and (b) the novel concept of merging called safe merge , which is a merge operation that we formally prove cannot cause a live HSDF graph to deadlock. The results show that the reduced graph (1) respects the throughput and latency constraints of the original application graph and (2) typically speeds up the process of extracting timing parameters and finding a feasible real-time schedule for real-time dataflow applications. They also show that when the throughput constraint is relaxed with respect to the maximal throughput of the graph, the merging algorithm is able to achieve a larger reduction in graph size, which in turn results in a larger speedup of the real-time scheduling algorithms.

Next generations of compute-intensive real-time applications in automotive systems will require more powerful computing platforms. One promising power-efficient solution for such applications is to use clustered many-core architectures. However, ensuring that real-time requirements are satisfied in the presence of contention in shared resources, such as memories, remains an open issue. This work presents a novel contention-free execution framework to execute automotive applications on such platforms. Privatization of memory banks together with defined access phases to shared memory resources is the backbone of the framework. An Integer Linear Programming (ILP) formulation is presented to find the optimal time-triggered schedule for the on-core execution as well as for the access to shared memory. Additionally a heuristic solution is presented that generates the schedule in a fraction of the time required by the ILP. Extensive evaluations show that the proposed heuristic performs only 0.5% away from the optimal solution while it outperforms a baseline heuristic by 67%. The applicability of the approach to industrially sized problems is demonstrated in a case study of a software for Engine Management Systems.

A task can be preempted by several jobs of higherpriority tasks during its response time. Assuming the worst-casememory demand for each of these jobs leads to pessimistic worst-case response time (WCRT) estimations. Indeed, there is a bigchance that a large portion of the instructions and data associatedwith the preempting task Tj are still available in the cache when Tj releases its next jobs. Accounting for this observation allowsthe pessimism of WCRT analysis to be significantly reduced, which is not considered by existing work. The four main contributions of this paper are: 1) The conceptof persistent cache blocks is introduced in the context of WCRTanalysis, which allows re-use of cache blocks to be captured,2) A cache-persistence-aware WCRT analysis for fixed-prioritypreemptive systems exploiting the PCBs to reduce the WCRTbound, 3) A multi-set extension of the analysis that furtherimproves the WCRT bound and 4) An evaluation showing thatour cache-persistence-aware WCRT analysis results in up to 10%higher schedulability than state-of-the-art approaches.

In modern multi-core systems with multiple real-time (RT) applications, memory traffic accessing the shared SDRAM is increasingly diverse, e.g., transactions have variable sizes. RT memory controllers with dynamic command scheduling can efficiently address the diversity by issuing appropriate commands subject to the SDRAM timing constraints. However, the scheduling dependencies between commands make it challenging to derive tight bounds for the worst-case response time (WCRT) and the worst-case bandwidth (WCBW) of a memory controller. Existing modeling and analysis techniques either do not provide tight WCRT and WCBW bounds for diverse memory traffic with variable transaction sizes or are difficult to adapt to different RT memory controllers. This paper models a memory controller using Timed Automata (TA), where model checking is applied for analysis. Our TA model is modular and accurately captures the behavior of a RT memory controller with dynamic command scheduling. We obtain WCRT and WCBW bounds, which are validated by simulating the worst- case transaction traces obtained by model checking with a cycle-accurate model of the memory controller. Our method outperforms three state-of-the-art analysis techniques. We reduce WCRT bound by up to 20%, while the average improvement is 7.7%, and increase the WCBW bound by up to 25% with an average improvement of 13.6%. In addition, our modeling is generic enough to extend to memory controllers with different mechanisms.

Memory controller design is challenging as mixed time-criticality embedded systems feature an increasing diversity of real-time (RT) and non-real-time (NRT) applications with variable transaction sizes. To satisfy the requirements of the applications, tight bounds on the worst-case response time (WCRT) of memory transactions must be provided to RT applications, while the lowest possible average response time must be given to the remaining applications. Existing real-time memory controllers cannot efficiently achieve this goal as they either bound the WCRT by sacrificing the average response time, or cannot efficiently support variable transaction sizes. In this article, we propose to use dynamic command scheduling, which is capable of efficiently dealing with transactions with variable sizes. The three main contributions of this article are: (1) a memory controller architecture consisting of a front-end and a back-end, where the former uses a TDM arbiter with a new work-conserving policy and the latter has a dynamic command scheduling algorithm that is independent of the front-end, (2) a formalization of the timings of the memory transactions for the proposed algorithm and architecture, and (3) an analysis of WCRT for transactions to capture the behavior of both the front-end and the back-end. This WCRT analysis supports variable transaction sizes and different degrees of bank parallelism. The critical part of the WCRT is the worst-case execution time (WCET) of a transaction, which is the time spent on command scheduling in the back-end. The WCET is bounded by two techniques applied to both fixed and variable transaction sizes, respectively. We experimentally evaluate the proposed memory controller and compare to an existing semi-static approach. The results demonstrate that dynamic command scheduling significantly outperforms the semi-static approach in the average case, while it performs equally well or better in the worst-case with only a few exceptions. The former reduces the average response time for NRT applications, and the latter pertains the WCRT for RT applications.

Real-time safety-critical systems should provide hard bounds on an applications' performance. SDRAM controllers used in this domain should therefore have a bounded worst-case bandwidth, response time, and power consumption. Existing works on real-time SDRAM controllers only consider a narrow range of memory devices, and do not evaluate how their schedulers' performance varies across memory generations, nor how the scheduling algorithm influences power usage. The extent to which the number of banks used in parallel to serve a request impacts performance is also unexplored, and hence there are gaps in the tool set of a memory subsystem designer, in terms of both performance analysis, and configuration options. This article introduces a generalized close-page memory command scheduling algorithm that uses a variable number of banks in parallel to serve a request. To reduce the schedule length for DDR4 memories, we exploit bank grouping through a pairwise bank-group interleaving scheme. The algorithm is evaluated using an ILP formulation, and provides schedules of optimal length for most of the considered LPDDR, DDR2, DDR3, LPDDR2, LPDDR3 and DDR4 devices. We derive the worst-case bandwidth, power and execution time for the same set of devices, and discuss the observed trade-offs and trends in the scheduler-configuration design space based on these metrics, across memory generations.

Many embedded multi-core systems incorporate both dataflow applications with timing constraints and traditional real-time applications. Applying real-time scheduling techniques on such systems provides real-time guarantees that all running applications will execute safely without violating their deadlines. However, to apply traditional real-time scheduling techniques on such mixed systems, a unified model to represent both types of applications running on the system is required. Several earlier works have addressed this problem and solutions have been proposed that address acyclic graphs, implicit-deadline models or are able to extract timing parameters considering specific scheduling algorithms. In this paper, we present an algorithm for extracting real-time parameters (offsets, deadlines and periods) that are independent of the schedulability analysis, other applications running in the system, and the specific platform. The proposed algorithm: 1) enables applying traditional real-time schedulers and analysis techniques on cyclic or acyclic Homogeneous Synchronous Dataflow (HSDF) applications with periodic sources, 2) captures overlapping iterations, which is a main characteristic of the execution of dataflow applications, 3) provides a method to assign offsets and individual deadlines for HSDF actors, and 4) is compatible with widely used deadline assignment techniques, such as NORM and PURE. The paper proves the correctness of the proposed algorithm through formal proofs and examples.

Complex contemporary systems contain multiple applications, some which have firm real-time requirements while others do not. These applications are deployed on multi-core platforms with shared resources, such as processors, interconnect, and memories. However, resource sharing causes contention between sharing applications that must be resolved by a resource arbiter. Time-Division Multiplexing (TDM) is a commonly used arbiter, but it is challenging to configure such that the bandwidth and latency requirements of the real-time resource clients are satisfied, while minimizing their total allocation to improve the performance of non-real-time clients. This work addresses this problem by presenting an efficient TDM configuration methodology. The five main contributions are: 1) An analysis to derive a bandwidth and latency guarantee for a TDM schedule with arbitrary slot assignment, 2) A formulation of the TDM configuration problem and a proof that it is NP-hard, 3) An integer-linear programming model that optimally solves the configuration problem by exhaustively evaluating all possible TDM schedule sizes, 4) A heuristic method to choose candidate schedule sizes that substantially reduces computation time with only a slight decrease in efficiency, 5) An experimental evaluation of the methodology that examines its scalability and quantifies the trade-off between computation time and total allocation for the optimal and the heuristic algorithms. The approach is also demonstrated on a case study of a HD video and graphics processing system, where a memory controller is shared by a number of processing elements.

Real-time systems need time-predictable platforms to allow static analysis of the worst-case execution time (WCET). Standard multi-core processors are optimized for the average case and are hardly analyzable. Within the T-CREST project we propose novel solutions for time-predictable multi-core architectures that are optimized for the WCET instead of the average-case execution time. The resulting time-predictable resources (processors, interconnect, memory arbiter, and memory controller) and tools (compiler, WCET analysis) are designed to ease WCET analysis and to optimize WCET performance. Compared to other processors the WCET performance is outstanding. The T-CREST platform is evaluated with two industrial use cases. An application from the avionic domain demonstrates that tasks executing on different cores do not interfere with respect to their WCET. A signal processing application from the railway domain shows that the WCET can be reduced for computation-intensive tasks when distributing the tasks on several cores and using the network-on-chip for communication. With three cores the WCET is improved by a factor of 1.8 and with 15 cores by a factor of 5.7. The T-CREST project is the result of a collaborative research and development project executed by eight partners from academia and industry. The European Commission funded T-CREST.

This article presents a semantics-based program verification framework for critical embedded real-time systems using the worst-case execution time (WCET) as the safety parameter. The verification algorithm is designed to run on devices with limited computational resources where efficient resource usage is a requirement. For this purpose, the framework of abstract-carrying code (ACC) is extended with an additional verification mechanism for linear programming (LP) by applying the certifying properties of duality theory to check the optimality of WCET estimates. Further, the WCET verification approach preserves feasibility and scalability when applied to multicore architectural models. The certifying WCET algorithm is targeted to architectural models based on the ARM instruction set and is presented as a particular instantiation of a compositional data-flow framework supported on the theoretic foundations of denotational semantics and abstract interpretation. The data-flow framework has algebraic properties that provide algorithmic transformations to increase verification efficiency, mainly in terms of verification time. The WCET analysis/verification on multicore architectures applies the formalism of latency-rate (LR) servers, and proves its correctness in the context of abstract interpretation, in order to ease WCET estimation of programs sharing resources.

Ever-increasing demands for main memory bandwidth and memory speed/power tradeoff led to the introduction of memories with multiple memory channels, such as Wide IO DRAM. Efficient utilization of a multichannel memory as a shared resource in multiprocessor real-time systems depends on mapping of the memory clients to the memory channels according to their requirements on latency, bandwidth, communication, and memory capacity. However, there is currently no real-time memory controller for multichannel memories, and there is no methodology to optimally configure multichannel memories in real-time systems. As a first work toward this direction, we present two main contributions in this article: (1) a configurable real-time multichannel memory controller architecture with a novel method for logical-to-physical address translation and (2) two design-time methods to map memory clients to the memory channels, one an optimal algorithm based on an integer programming formulation of the mapping problem, and the other a fast heuristic algorithm. We demonstrate the real-time guarantees on bandwidth and latency provided by our multichannel memory controller architecture by experimental evaluation. Furthermore, we compare the performance of the mapping problem formulation in a solver and the heuristic algorithm against two existing mapping algorithms in terms of computation time and mapping success ratio. We show that an optimal solution can be found in 2 hours using the solver and in less than 1 second with less than 7% mapping failure using the heuristic for realistically sized problems. Finally, we demonstrate configuring a Wide IO DRAM in a high-definition (HD) video and graphics processing system to emphasize the practical applicability and effectiveness of this work.

Scaling CMOS technology into nanometer feature-size nodes has made it practically impossible to precisely control the manufacturing process. This results in variation in the speed and power consumption of a circuit. As a solution to process-induced variations, circuits are conventionally implemented with conservative design margins to guarantee the target frequency of each hardware component in manufactured multiprocessor chips. This approach, referred to as worst-case design, results in a considerable circuit upsizing, in turn reducing the number of dies on a wafer. This work deals with the design of real-time systems for streaming applications (e.g., video decoders) constrained by a throughput requirement (e.g., frames per second) with reduced design margins, referred to as better-than-worst-case design . To this end, the first contribution of this work is a complete modeling framework that captures a streaming application mapped to an NoC-based multiprocessor system with voltage-frequency islands under process-induced die-to-die and within-die frequency variations . The framework is used to analyze the impact of variations in the frequency of hardware components on application throughput at the system level . The second contribution of this work is a methodology to use the proposed framework and estimate the impact of reducing circuit design margins on the number of good dies that satisfy the throughput requirement of a real-time streaming application . We show on both synthetic and real applications that the proposed better-than-worst-case design approach can increase the number of good dies by up to 9.6% and 18.8% for designs with and without fixed SRAM and IO blocks, respectively.

Manufacturing-time process (P) variations and runtime voltage (V) and temperature (T) variations can affect a DRAM's performance severely. To counter these effects, DRAM vendors provide substantial design-time PVT timing margins to guarantee correct DRAM functionality under worst-case operating conditions. Unfortunately, with technology scaling these timing margins have become large and very pessimistic for a majority of the manufactured DRAMs. While run-time variations are specific to operating conditions and as a result, their margins difficult to optimize, process variations are manufacturing-time effects and excessive process-margins can be reduced at run-time, on a per-device basis, if properly identified. In this paper, we propose a generic post-manufacturing performance characterization methodology for DRAMs that identifies this excess in process-margins for any given DRAM device at runtime, while retaining the requisite margins for voltage (noise) and temperature variations. By doing so, the methodology ascertains the actual impact of process-variations on the particular DRAM device and optimizes its access latencies (timings), thereby improving its overall performance. We evaluate this methodology on 48 DDR3 devices (from 12 DIMMs) and verify the derived timings under worst-case operating conditions, showing up to 33.3% and 25.9% reduction in DRAM read and write latencies, respectively.

Existing memory subsystems and TDM NoCs for real-time systems are optimized independently in terms of cost and performance by configuring their arbiters according to the bandwidth and/or latency requirements of their clients. However, when they are used in conjunction, and run in different clock domains, i.e. they are decoupled, there exists no structured methodology to select the NoC interface width and operating frequency for minimizing area and/or power consumption. Moreover, the multiple arbitration points, one in the NoC and the other in the memory subsystem, introduce additional overhead in the worst-case guaranteed latency. These makes it hard to design cost-efficient real-time systems. The three main contributions in this paper are: (1) We present a novel methodology to couple any existing TDM NoC with a realtime memory controller and compute the different NoC interface width and operating frequency combinations for minimal area and/or power consumption. (2) For two different TDM NoC types, one a packet-switched and the other circuit-switched, we show the trade-off between area and power consumption with the different NoC configurations, for different DRAM generations. (3) We compare the coupled and decoupled architectures with the two NoCs, in terms of guaranteed worst-case latency, area and power consumption by synthesizing the designs in 40 nm technology. Our experiments show that using a coupled architecture in a system consisting of 16 clients results in savings of over 44% in guaranteed latency, 18% and 17% in area, 19% and 11% in power consumption for a packet-switched and a circuit-switched TDM NoC, respectively, with different DRAM types.

Memory controller design is challenging as real-time embedded systems feature an increasing diversity of real-time and non-real-time applications with variable transaction sizes. To satisfy the requirements of the applications, tight bounds on the worst-case execution time (WCET) of memory transactions must be provided to real-time applications, while the lowest possible average execution time must be given to the rest. Existing real-time memory controllers cannot efficiently achieve this goal as they either bound the WCET by sacrificing the average execution time, or are not scalable to directly support variable transaction sizes, or both. In this paper, we propose to use dynamic command scheduling, which is capable of efficiently dealing with transactions with variable sizes. The three main contributions of this paper are: 1) a back-end architecture for a real-time memory controller with a dynamic command scheduling algorithm, 2) a formalization of the timings of the memory transactions for the proposed architecture and algorithm, and 3) two techniques to bound the WCET of transactions with both fixed and variable sizes, respectively. We experimentally evaluate the proposed memory controller and compare both the worst-case and average-case execution times of transactions to a state-of-the-art semi-static approach. The results demonstrate that dynamic command scheduling outperforms the semi-static approach by 33.4% in the average case and performs at least equally well in the worst case. We also show the WCET is tight for transactions with fixed and variable sizes, respectively.

As technology scales, the impact of process variation on the maximum supported frequency (FMAX) of individual cores in a multiprocessor system-on-chip (MPSoC) becomes more pronounced. Task allocation without variation-aware performance analysis can greatly compromise performance and lead to a significant loss in yield , defined as the percentage of manufactured chips satisfying the application timing requirement . We propose variation-aware task allocation for best-effort and real-time streaming applications modeled as task graphs. Our solutions are primarily based on the throughput requirement, which is the most important timing requirement in many real-time streaming applications. The four main contributions of this work are (1) distinguishing best-effort firm real-time and soft real-time application classes, which require different optimization criteria, (2) using dataflow graphs, which are well suited for modeling and analysis of streaming applications, we explicitly model task execution both in terms of clock cycles (which is independent of variation) and seconds (which does depend on the variation of the resource), which we connect by an explicit binding, (3) we present two optimization approaches, which give different improvement results at different costs, (4) we present both exhaustive and heuristic algorithms that implement the optimization approaches. Our variation-aware mapping algorithms are tested on models of seven real applications and are compared to mapping methods that are unaware of hardware variation. Our results demonstrate (1) improvements in the average performance (3% on average) for best-effort applications, and (2) for firm real-time and soft real-time applications, yield improvements of up to 27% with an average of 15%, showing the effectiveness of our approaches.

DRAM vendors provide pessimistic current measures in memory datasheets to account for worst-case impact of process variations and to improve their production yield, leading to unrealistic power consumption estimates. In this paper, we first demonstrate the possible effects of process variations on DRAM performance and power consumption by performing Monte-Carlo simulations on a detailed DRAM cross-section. We then propose a methodology to empirically determine the actual impact for any given DRAM memory by assessing its performance characteristics during the DRAM calibration phase at system boot-time, thereby enabling its optimal use at run-time. We further employ our analysis on Micron's 2Gb DDR3-1600-x16 memory and show considerable over-estimation in the datasheet measures and the energy estimates (up to 28%), by using realistic current measures for a set of MediaBench applications.

JEDEC recently introduced its new standard for 3D-stacked Wide I/O DRAM memories, which defines their architecture, design, features and timing behavior. With improved performance/power trade-offs over previous generation DRAMs, Wide I/O DRAMs provide an extremely energy-efficient green memory solution required for next-generation embedded and high-performance computing systems. With both industry and academia pushing to evaluate and employ these highly anticipated memories, there is an urgent need for an accurate power model targeting Wide I/O DRAMs that enables their efficient integration and energy management in DRAM stacked SoC architectures. In this paper, we present the first system-level power model of 3D-stacked Wide I/O DRAM memories that is almost as accurate as detailed circuit-level power models of 3D-DRAMs. To verify its accuracy, we experimentally compare its power and energy estimates for different memory workloads and operations against those of a circuit-level 3D-DRAM power model and show less than 2% difference between the two sets of estimates.

Optimal utilization of a multi-channel memory, such as Wide IO DRAM, as shared memory in multi-processor platforms depends on the mapping of memory clients to the memory channels, the granularity at which the memory requests are interleaved in each channel, and the bandwidth and memory capacity allocated to each memory client in each channel. Firm real-time applications in such platforms impose strict requirements on shared memory bandwidth and latency, which must be guaranteed at design-time to reduce verification effort. However, there is currently no real-time memory controller for multichannel memories, and there is no methodology to optimally configure multi-channel memories in real-time systems. This paper has four key contributions: (1) A real-time multi-channel memory controller architecture with a new programmable Multi-Channel Interleaver unit. (2) A novel method for logical-to-physical address translation that enables inter-leaving memory requests across multiple memory channels at different granularities. (3) An optimal algorithm based on an Integer Linear Program (ILP) formulation to map memory clients to memory channels considering their communication dependencies, and to configure the memory controller for minimum bandwidth utilization. (4) We experimentally evaluate the run-time of the algorithm and show that an optimal solution can be found within 15 minutes for realistically sized problems. We also demonstrate configuring a multi-channel Wide IO DRAM in a High-Definition (HD) video and graphics processing system to emphasize the effectiveness of our approach.

Complex Systems-on-Chips (SoC) are mixed time-criticality systems that have to support firm real-time (FRT) and soft real-time (SRT) applications running in parallel. This is challenging for critical SoC components, such as memory controllers. Existing memory controllers focus on either firm real-time or soft real-time applications. FRT controllers use a close-page policy that maximizes worst-case performance and ignore opportunities to exploit locality, since it cannot be guaranteed. Conversely, SRT controllers try to reduce latency and consequently processor stalling by speculating on locality. They often use an open-page policy that sacrifices guaranteed performance, but is beneficial in the average case. This paper proposes a conservative open-page policy that improves average-case performance of a FRT controller in terms of bandwidth and latency without sacrificing real-time guarantees. As a result, the memory controller efficiently handles both FRT and SRT applications. The policy keeps pages open as long as possible without sacrificing guarantees and captures locality in this window. Experimental results show that on average 70% of the locality is captured for applications in the CHStone benchmark, reducing the execution time by 17% compared to a close-page policy. The effectiveness of the policy is also evaluated in a multi-application use-case, and we show that the overall average-case performance improves if there is at least one FRT or SRT application that exploits locality.

The transition towards multi-processor systems with shared resources is challenging for real-time systems, since resource interference between concurrent applications must be bounded using timing analysis. There are two common approaches to this problem: 1) Detailed analysis that models the particular resource and arbiter cycle-accurately to achieve tight bounds. 2) Using temporal abstractions, such as latency-rate (LR) servers, to enable unified analysis for different resources and arbiters using well-known timing analysis frameworks. However, the use of abstraction typically implies reducing the tightness of analysis that may result in over-dimensioned systems, although this pessimism has not been properly investigated. This paper compares the two approaches in terms of worst-case execution time (WCET) of applications sharing an SDRAM memory under Credit-Controlled Static-Priority (CCSP) arbitration. The three main contributions are: 1) A detailed interference analysis of the SDRAM memory and CCSP arbiter. 2) Based on the detailed analysis, two optimizations are proposed to the LR analysis that increase the tightness of its interference bounds. 3) An experimental comparison of the two approaches that quantifies their impact on the WCET of applications from the CHStone benchmark.

Contemporary embedded systems are based on complex heterogeneous multi-core platforms to cater to the increasing number of applications, some of which have (soft) real-time requirements. To reduce cost, resources are shared using diverse arbitration mechanisms, such as Time-Division Multiplexing (TDM), Static-Priority (SP), and Round-Robin (RR), depending on application and resource requirements. However, resource sharing results in interference between sharing applications making it difficult to estimate if the average latency is sufficient to satisfy their real-time requirements. Existing work proposes isolated models that either fail to address the diversity of arbitration mechanisms or cannot capture the dynamic arrival and service processes of applications and resources in multi-core platforms. This paper addresses this problem by proposing a general framework for average-case performance analysis of shared resources in multi-core platforms. The two main contributions are: 1) a general model for resource sharing based on queuing theory that can be used with different arbiters and that captures architectural features of the shared resource, such as pipelining and arbitration delay, and 2) three arbiter models for TDM, SP, and RR, respectively that assume general distributions (G/G/1) and fits within the framework.

Designing cost-efficient multi-core real-time systems requires efficient techniques to allocate applications to cores while satisfying their timing constraints. However, existing approaches typically allocate using a First-Fit algorithm, which does not consider the execution time and potential parallelism of paths in the applications, resulting in over-dimensioned systems. This work addresses this problem by proposing a new heuristic algorithm, Critical-Path-First, for the allocation of real-time streaming applications modeled as dataflow graphs on 2D mesh multi-core processors. The main criteria of the algorithm is to allocate paths that have the highest impact on the execution time of the application first. It is also able to exploit parallelism in the application by allocating parallel paths on different cores. Experimental evaluation shows that the proposed heuristic improves the resource utilization by allocating up to 7% more applications and it minimizes the average end-to-end worst-case response time of the allocated applications by up to 31%.

Powering down SDRAMs at run-time reduces memory energy consumption significantly, but often at the cost of performance. If employed speculatively with real-time memory controllers, power-down mechanisms could impact both the guaranteed bandwidth and the memory latency bounds. This calls for power-down strategies that can hide or bound the performance loss, making run-time memory power-down feasible for real-time applications.

The performance and power consumption of mobile DRAMs (LPDDRs) depend on the configuration of system-level parameters, such as operating frequency, interface width, request size, and memory map. In mobile systems running both real-time and non-real-time applications, the memory configuration must satisfy bandwidth requirements of real-time applications, meet the power consumption budget, and offer the best average-case execution time to the non-real-time applications. There is currently no well-defined methodology for selecting a suitable memory configuration for real-time mobile systems. The worst-case bandwidth, average-case execution time, and power consumption of mobile DRAMs across generations have furthermore not been investigated. This paper has two main contributions. 1) We analyze the worst-case bandwidth, average-case execution time, and power consumption of mobile DRAMs across three generations: LPDDR, LPDDR2 and Wide-IO-based 3D-stacked DRAM. 2) Based on our analysis, we propose a methodology for selecting memory configurations in real-time mobile systems.We show that LPDDR (32-bit IO), LPDDR2 (32-bit IO) and 3D-DRAM (128-bit IO) provide worst-case bandwidth up to 0.75 GB/s, 1.6 GB/s and 3.1 GB/s, respectively. We furthermore show for an H.263 decoder that LPDDR2 and 3D-DRAM reduce power consumption with up to 25% and 67%, respectively, compared to LPDDR, and reduce the execution time with up to 18% and 25%.

A modern real-time embedded system must support multiple concurrently running applications. To reduce costs, critical SoC components like SDRAM memories are often shared between applications with a variety of firm real-time requirements. To guarantee that the system works as intended, the memory controller must be configured such that all the real-time requirements of all sharing applications are satisfied. The attainable worst-case bandwidth, latency, and power of the memory depend largely on memory map configuration. Sharing SDRAM amongst multiple applications is challenging, since their requirements might call for different memory maps. This paper presents an exploration of the memory-map design space. Two contributions improve the memory-map selection procedure. The first contribution reduces the minimum access granularity by interleaving requests over a configurable number of banks instead of all banks. This technique is beneficial for worst-case performance in terms of bandwidth, latency and power. As a second contribution, we present a methodology to derive a memory-map configuration, i.e. the access granularity and number of interleaved banks, from a specification of the real-time application requirements and an overall memory power budget.

Reducing power/energy consumption is an important goal for all computer systems, from servers to battery-driven hand-held devices. To achieve this goal, the energy consumption of all system components needs to be reduced. One of the most power-hungry components is the off-chip DRAM, even when it is idle. DRAMs support different power-saving modes, such as self-refresh and power-down, but employing them every time the DRAM is idle, reduces performance due to their power-up latencies. The self-refresh mode offers large power savings, but incurs a long power-up latency. The power-down mode, on the other hand, has a shorter power-up latency, but provides lower power savings. In this paper, we propose and evaluate a novel power-saving policy that combines the best of both power-saving modes in order to achieve significant power reductions with a marginal performance penalty. To accomplish this, we use a history-based predictor to forecast the duration of an idle period and then either employ self-refresh, or power-down, or a combination of both power saving modes. Significant refinements are made to the predictor to maximize the energy savings and minimize the performance penalty. The presented policy is evaluated using several applications from the multimedia domain and the experimental results show that it reduces the total DRAM energy consumption between 68.8% and 79.9% at a negligible performance penalty between 0.3% and 2.2%.

Designing multi-processor systems-on-chips becomes increasingly complex, as more applications with realtime requirements execute in parallel. System resources, such as memories, are shared between applications to reduce cost, causing their timing behavior to become inter-dependent. Using conventional simulation-based verification, this requires all concurrently executing applications to be verified together, resulting in a rapidly increasing verification complexity. Predictable and composable systems have been proposed to address this problem. Predictable systems provide bounds on performance, enabling formal analysis to be used as an alternative to simulation. Composable systems isolate applications, enabling them to be verified independently. Predictable and composable systems are built from predictable and composable resources. This paper presents three general techniques to implement and model predictable and composable resources, and demonstrates their applicability in the context of a memory controller. The architecture of the memory controller is general and supports both SRAM and DDR2/DDR3 SDRAM and a wide range of arbiters, making it suitable for many predictable and composable systems. The modeling approach is based on a shared-resource abstraction that covers any combination of supported memory and arbiter and enables system-level performance analysis with a variety of well-known frameworks, such as network calculus or data-flow analysis.

Power modeling and estimation has become one of the most defining aspects in designing modern embedded systems. In this context, DDR SDRAM memories contribute significantly to system power consumption, but lack accurate and generic power models. The most popular SDRAM power model provided by Micron, is found to be inaccurate or insufficient for several reasons. First, it does not consider the power consumed when transitioning to power-down and self-refresh modes. Second, it employs the minimal timing constraints between commands from the SDRAM datasheets and not the actual duration between the commands as issued by an SDRAM memory controller. Finally, without adaptations, it can only be applied to a memory controller that employs a close-page policy and accesses a single SDRAM bank at a time. These critical issues with Micron's power model impact the accuracy and the validity of the power values reported by it and resolving them, forms the focus of our work. In this paper, we propose an improved SDRAM power model that estimates power consumption during the state transitions to power-saving states, employs an SDRAM command trace to get the actual timings between the commands issued and is generic and applicable to all DDRx SDRAMs and all memory controller policies and all degrees of bank interleaving. We quantitatively compare the proposed model against the unmodified Micron model on power and energy for DDR3-800. We show differences of up to 60% in energy-savings for the precharge power-down mode for a power-down duration of 14 cycles and up to 80% for the self-refresh mode for a self-refresh duration of 560 cycles.

Verifying firm real-time requirements gets increasingly complex, as the number of applications in embedded systems grows. Predictable systems reduce the complexity by enabling formal verification. However, these systems require predictable software and hardware components, which is problematic for resources with highly variable execution times, such as SDRAM controllers. A predictable SDRAM controller has been proposed that addresses this problem using predictable memory patterns, which are precomputed sequences of SDRAM commands. However, the memory patterns are derived manually, which is a time-consuming and error-prone process that must be repeated for every memory device, and may result in inefficient use of scarce and expensive bandwidth. This paper addresses this issue by proposing three algorithms for automatic generation of efficient memory patterns that provide different trade-offs between run-time of the algorithm and the bandwidth guaranteed by the controller. We experimentally evaluate the algorithms for a number of DDR2/DDR3 memories and show that an appropriate choice of algorithm reduces run-time to less than a second and increases the guaranteed bandwidth by up to 10.2%.

A present-day System-on-Chip (SoC) runs a wide range of applications with diverse real-time requirements. Resources, such as processors, interconnects and memories, are shared between these applications to reduce cost. Resource sharing causes temporal interference, which must be bounded by a suitable resource arbiter. System-level analysis techniques use the service guarantee of the arbiter to ensure that real-time requirements of these applications are satisfied. A service guarantee that underestimates the minimum service provided by an arbiter results in more allocation of resources than needed to satisfy latency and throughput requirements. For instance, a linear service guarantee cannot accurately capture burst service provision by many priority-based schedulers, such as Credit-Controlled Static Priority (CCSP) and Priority-Budget Scheduling (PBS). As a result, the timing analysis of these arbiters becomes too pessimistic. This leads to unnecessary cost penalties since some SoC resources, such as SDRAM bandwidth, are scarce and expensive. This paper addresses this problem for the CCSP arbiter. The two main contributions are: (1) a piecewise linear service guarantee that accurately captures bursty service provisioning, and (2) an equivalent dataflow model of the new service guarantee, which is an essential component to integrate the arbiter with dataflow-based system-level design techniques that analyze the worst-case latency and throughput of real-time applications. The new service guarantee enables efficient resource utilization under CCSP arbitration. Experimental results of an H.263 video decoder application show that memory bandwidth savings from 26% up to 67% can be achieved by using the new service guarantee as compared to the existing linear service guarantee.

The verification complexity of real-time requirements in embedded systems grows exponentially with the number of applications, as resource sharing prevents independent verification using simulation-based approaches. Formal verification is a promising alternative, although its applicability is limited to systems with predictable hardware and software. SDRAM memories are common examples of essential hardware components with unpredictable timing behavior, typically preventing use of formal approaches. A predictable SDRAM controller has been proposed that provides guarantees on bandwidth and latency by dynamically scheduling memory patterns, which are statically computed sequences of SDRAM commands. However, the proposed patterns become increasingly inefficient as memories become faster, making them unsuitable for DDR3 SDRAM. This paper extends the memory pattern concept in two ways. Firstly, we introduce a burst count parameter that enables patterns to have multiple SDRAM bursts per bank, which is required for DDR3 memories to be used efficiently. Secondly, we present a classification of memory pattern sets into four categories based on the combination of patterns that cause worst-case bandwidth and latency to be provided. Bounds on bandwidth and latency are derived that apply to all pattern types and burst counts, as opposed to the single case covered by earlier work. Experimental results show that these extensions are required to support the most efficient pattern sets for many use-cases. We also demonstrate that the burst count parameter increases efficiency in presence of large requests and enables a wider range of real-time requirements to be satisfied.

Verification of application requirements is becoming a bottleneck in system-on-chip design, as the number of applications grows. Traditionally, the verification complexity increases exponentially with the number of applications and must be repeated if an application is added, removed, or modified. Predictable systems offering lower bounds on performance have been proposed to manage the increasing verification complexity, although this approach is only applicable to a restricted set of applications and systems. Composable systems, on the other hand, completely isolate applications in both the value and time domains, allowing them to be independently verified. However, existing approaches to composable system design are either restricted to applications that can be statically scheduled, or share resources using time-division multiplexing, which cannot efficiently satisfy tight latency requirements. In this paper, we present an approach to composable resource sharing based on latency-rate servers that supports any arbiter belonging to the class, providing a larger solution space for a given set of requirements. The approach can be combined with formal performance analysis using a variety of well-known modeling frame works. We furthermore propose an architecture for a resource front end that implements our concepts and provides composable service for any resource with bounded service time. The architecture supports both systems with buffers dimensioned to prevent overflow and systems with smaller buffers, where overflow is prevented with flow control. Finally, we experimentally demonstrate the usefulness of our approach with a simple use case sharing an SRAM memory.

Resources in contemporary systems-on-chip (SoC) are shared between applications to reduce cost. Access to shared resources is provided by arbiters that require a small hardware implementation and must run at high speed. To manage heavily loaded resources, such as memory channels, it is also important that the arbiter minimizes over allocation. A Credit-Controlled Static-Priority (CCSP) arbiter comprised of a rate regulator and a static-priority scheduler has been proposed for scheduling access to SoC resources. The proposed rate regulator, however, is not straight-forward to implement in hardware, and assumes that service is allocated with infinite precision. In this paper, we introduce a fast and small hardware implementation of the CCSP rate regulator and formally prove its correctness. We also show an efficient way of representing the allocated service in hardware with finite precision. Based on this representation, we define and evaluate two allocation strategies, and derive tight bounds on their respective over allocations. We show that increasing the precision of the implementation results in an exponential reduction in maximum over allocation at the cost of a linear increase in area. We demonstrate that the allocation strategy has a large impact on the allocation success rate for use cases with high load. Finally, we compare CCSP to traditional frame-based approaches and conclude that having a fine allocation granularity that is decoupled from latency is essential to manage highly loaded resources in real-time systems.

The convergence of application domains in new systems-on-chip (SoC) results in systems with many applications with a mix of soft and hard real-time requirements. To reduce cost, resources, such as memories and interconnect, are shared between applications. However, resource sharing introduces interference between the sharing applications, making it difficult to satisfy their real-time requirements. Existing arbiters do not efficiently satisfy the requirements of applications in SoCs, as they either couple rate or allocation granularity to latency, or cannot run at high speeds in hardware with a low-cost implementation. The contribution of this paper is an arbiter called credit- controlled static-priority (CCSP), consisting of a rate regulator and a static-priority scheduler. The rate regulator isolates applications by regulating the amount of provided service in a way that decouples allocation granularity and latency. The static-priority scheduler decouples latency and rate, such that low latency can be provided to any application, regardless of the allocated rate. We show that CCSP belongs to the class of latency-rate servers and guarantees the allocated rate within a maximum latency, as required by hard real-time applications. We present a hardware implementation of the arbiter in the context of a DDR2 SDRAM controller. An instance with six ports running at 200 MHz requires an area of 0.0223 mm2in a 90 nm CMOS process.