Decoupled Access-Execute (DAE) architectures separate memory accesses from computation in two specialized units. This design is becoming increasingly popular among hyperscalers to accelerate irregular embedding lookups in recommendation models. In this paper, we first broaden the scope by demonstrating the benefits of DAE architectures across a wider range of irregular embedding operations in several machine learning models. Then, we propose the Ember compiler to automatically compile all of these embedding operations to DAE architectures. Conversely from other DAE compilers, Ember features multiple intermediate representations specifically designed for different optimization levels. In this way, Ember can implement all optimizations to match the performance of hand-written code, unlocking the full potential of DAE architectures at scale.

Current heterogeneous systems integrate generalpurpose Central Processing Units (CPUs), Graphics Processing Units (GPUs), and Neural Processing Units (NPUs). The efficient use of such systems requires a significant programming effort to distribute computation and synchronize across devices, which usually involves using Atomic Memory Operations (AMOs). Arm recently launched a floating-point Atomic Memory Operations (FAMOs) extension to perform atomic updates on floating-point data types specifically. This work characterizes and models heterogeneous architectures to understand how floating-point AMOs impact graph, Machine Learning (ML), and high-performance computing (HPC) workloads. Our analysis shows that many AMOs are performed on floating-point data, which modern systems execute using inefficient compare-and-swap (CAS) constructs. Therefore, replacing CASbased constructs with FAMOs can improve a wide range of workloads. Moreover, we analyze the trade-offs of executing FAMOs at different memory hierarchy levels, either in private caches (near) or remotely in shared caches (far). We have extended the widely used AMBA CHI protocol to evaluate such FAMO support on a simulated chiplet-based heterogeneous architecture. While near FAMOs achieve an average $1.34 \times$ speed-up, far FAMOs reach an average $1.58 \times$ speed-up. We conclude that FAMOs can bridge the gap between CPU architecture and accelerators and enabling synchronization in key application domains.

During the last decades, critical embedded systems in many application domains have transitioned from federated architectures to integrated architectures like Integrated Modular Avionics (IMA) platforms. Such trend is driven by the objective to reduce the size and weight of on-board equipment, the dissipated power and the number of computing unit types (and so to improve the maintainability and serviceability of the equipment). This evolution along with the growing need for more functionalities push toward the use of high-performance processors. Therefore, Commercial Off-The-Shelf (COTS) processors are now commonly used in many safety-critical systems to benefit from the huge processing power and the low cost of components designed for high-volume markets. But, ensuring that these components developed for mainstream markets can safely be used in critical applications is critical as their failure could jeopardize the success of the mission or the safety of persons. Most of the mission-critical systems have to meet specific and stringent non-functional requirements. They notably have to operate in harsh environmental conditions and to respect hard real-time constraints. Consequently, the development of embedded systems for application domains like avionics, space or transportation systems is primarily driven by their safety and reliability requirements. A structured and mastered development process following safety standards aims to guarantee that all the nonfunctional requirements are met. The traceability of the system requirements from the specifications to the final system is thus mandatory. To prevent costly late design changes (that could compromise the whole product development), verification steps have to be followed at each stage of the development from early system models to final HW prototypes. But, existing development processes and methods are now challenged by the increasing integration of mission-critical systems. With the shift to integrated computing architectures, multiple applications that were executed on independent computing units in the past are now executed on the top of a common computing platform. For such platform, supporting the incremental certification of applications is a key point for an efficient development process of the applications. So, a strict time and space partitioning between these applications is ensured by the platform to guarantee the safety of the system and to enable a composability of the applications. In this context, the move to multi-core processors represents a challenge as the shared resources result in inter-tasks interferences, which are difficult to evaluate. Methodologies to evaluate and control these timing interferences are thus essential to enable the avionics domain to shift to the multicore paradigm. And with the increasing reliability concerns of advanced semiconductor technologies, the early reliability estimation of the system also becomes a challenge. The talk will present the requirements of mission-critical systems. The industrial challenges of their design and of their early prototyping are discussed.

Advanced multifunctional computing systems realized in forthcoming technologies hold the promise of a significant increase of the computational capability that will offer end-users ever improving services and functionalities (e.g., next generation mobile devices, cloud services, etc.). However, the same path that is leading technologies toward these remarkable achievements is also making electronic devices increasingly unreliable, posing a threat to our society that is depending on the ICT in every aspect of human activities. Reliability of electronic systems is therefore a key challenge for the whole ICT technology and must be guaranteed without penalizing or slowing down the characteristics of the final products. CLERECO EU FP7 (GA No. 611404) research project addresses early accurate reliability evaluation and efficient exploitation of reliability at different design phases, since these aspects are two of the most important and challenging tasks toward this goal.

Evaluation of computing systems reliability must be accurate enough to provide hints for the required fault protection mechanisms that will guarantee correctness of operation at acceptance costs. To be useful, reliability evaluation must be performed early enough in the design cycle when, however, the available details of the system are largely unknown. This inherent contradiction in terms: early vs. accurate, requires a cross-layer approach for reliability evaluation. Different layers of abstraction contribute differently in the overall system reliability; if this contribution can be assessed independently, the reliability of the system can be evaluated at the early stages of the design. We review the state-of-the-art in the area and discuss corresponding challenges .

The computing market has been dominated during the last two decades by the well-known convergence of the high-performance computing market and the mobile market. In this paper we witness a new type of convergence between the mission-critical market (such as avionic or automotive) and the mainstream consumer electronics market. Such convergence is fuelled by the common needs of both markets for more reliability, support for mission-critical functionalities and the challenge of harnessing the unsustainable increases in safety margins to guarantee either correctness or timing. In this position paper, we present a description of this new convergence, as well as the main challenges and opportunities that it brings to computing industry.

Forthcoming technology nodes are posing major challenges on the manufacturing of reliable (real-time) systems: process variations, accelerated degradation aging, as well as external and internal noise are key examples. This paper focuses on real-time systems reliability and analyzes the state-of-the-art and the emerging reliability bottlenecks from three different perspectives: technology, circuit/IP and full system.

Summary form only given. The major bottleneck for technology scaling is the growing rate of hardware failures. Process variations are becoming extreme and sensitivity to radiation is becoming severe. In addition, intrinsic failures such as device parameter degradation are accelerating the wear-out. All of these are leading to higher random in-filed failures and shorter device lifetime. The 2011 ITRS (International Technology Roadmap for Semiconductors) projects very high bit failure rates of the order of 10-2for SRAM and of 10-3for latches for 16nm high performance technology. Hence, solving reliability challenges for future technologies requires new efficient and cost effective approaches not only to detect and recover from in-filed failures, but also to extend the device lifetime for targeted applications.

Recently, system designers are facing the challenge of developing systems that have diverse features, are more complex and more powerful, with less power consumption and reduced time to market. These contradictory constraints have forced technology providers to pursue design solutions that will allow design teams to meet the above design targets. In that respect, this paper introduces an innovative technology platform, called MORPHEUS, which intents to provide complete design framework for dealing with the aforementioned challenges. MORPHEUS consists of a state of the art architecture that encompasses heterogeneous reconfigurable accelerators for implementing on the same hardware architecture applications with varying characteristics and a tool chain that, through a software oriented approach, eases the implementation of highly complex applications with heterogeneous characteristics. The proposed approach has been tested and evaluated through state of the art cases studies borrowed from complementary application domains.

Commercial Off-The-Shelf (COTS) processors are now commonly used in real-time embedded systems. The characteristics of these processors fulfill system requirements in terms of time-to-market, low cost, and high performance-per-watt ratio. However, multithreaded (MT) processors are still not widely used in real-time systems because the timing analysis is too complex. In MT processors, simultaneously-running tasks share and compete for processor resources, so the timing analysis has to estimate the possible impact that the inter-task interferences have on the execution time of the applications. In this paper, we propose a method that quantifies the slowdown that simultaneously-running tasks may experience due to collision in shared processor resources. To that end, we designed benchmarks that stress specific processor resources and we used them to (1) estimate the upper limit of a slowdown that simultaneously-running tasks may experience because of collision in different shared processor resources, and (2) quantify the sensitivity of time-critical applications to collision in these resources. We used the presented method to determine if a given MT processor is a good candidate for systems with timing requirements. We also present a case study in which the method is used to analyze three multithreaded architectures exhibiting different configurations of resource sharing. Finally, we show that measuring the slowdown that real applications experience when simultaneously-running with resource-stressing benchmarks is an important step in measurement-based timing analysis. This information is a base for incremental verification of MT COTS architectures.

Process scaling has given designers billions of transistors to work with. As feature sizes near the atomic scale, extensive variation and wear-out inevitably make margining uneconomical or impossible. In this context, new design approaches are required. The inherent regularity and redundancy of SIMD architectures make them suitable to address the challenges posed by new semiconductor technologies at the architecture level. This paper proposes a built-in self-test/self-diagnosis procedure for a SIMD processor. Concurrent BIST operations are carried out after reset at each PE, thus resulting in scalable test application time with processor size. The key principle consists of exploiting the inherent structural redundancy of the SIMD architecture in a cooperative way, thus strongly reducing the testing framework latency and area overhead. Once the faults are detected, a reconfiguration technique is then proposed in order to preserve correct operation. Testing of single stuck-at faults is performed at-speed in 240 cycles regardless of the accelerator size, with a hardware overhead of less than 10%. Finally, the fault-tolerant tile integrating both BIST, reconfiguration logic and spare PE requires a 25% of total area overhead.

Performance demand of Critical Real-Time Embedded (CRTE) systems implementing safety-related system features grows at an exponential rate. Only modern semiconductor technologies can satisfy CRTE systems performance needs efficiently. However, those technologies lead to high failure rates, thus lowering survivability of chips to unacceptable levels for CRTE systems. This paper presents SESACS architecture (Surviving Errors in SAfety-Critical Systems), a paradigm shift in the design of CRTE systems. SESACS is a new system design methodology consisting of three main components: (i) a multicore hardware/firmware platform capable of detecting and diagnosing hardware faults of any type with minimal impact on the worst-case execution time (WCET), recovering quickly from errors, and properly reconfiguring the system so that the resulting system exhibits a predictable and analyzable degradation in WCET; (ii) a set of analysis methods and tools to prove the timing correctness of the reconfigured system; and (iii) a white-box methodology and tools to prove the functional safety of the system and compliance with industry standards. This new design paradigm will deliver huge benefits to the embedded systems industry for several decades by enabling the use of more cost-effective multicore hardware platforms built on top of modern semiconductor technologies, thereby enabling higher performance, and reducing weight and power dissipation. This new paradigm will further extend the life of embedded systems, therefore, reducing warranty and early replacement costs.

Reconfigurable architectures and NoC (Network-on-Chip) have introduced new research directions for technology and flexibility issues, which have been largely investigated in the last decades. Exploiting run-time adaptivity opens a new area of research by considering dynamic reconfiguration. In this paper, we present the architecture and associated development tools of an heterogeneous reconfigurable SoC focusing on the chosen communication infrastructure. The SOC integrates units of various sizes of reconfiguration granularity. The included NoC approach demonstrates the mentioned benefits and scalability for actual and future SoC design. On a reference CMOS090 implementation the described interconnect system works at the system reference frequency of 200 MHZ sustaining the required run-time bandwidth on a set of reference applications, at a price ≪ 10% in area in power consumption with respect to the overall system.