Search

With the prevalence of hardware accelerators as an integral part of the modern systems on chip (SoCs), the ability to model accelerators quickly and accurately within the system in which it operates is critical. This paper presents gem5-SALAMv2 as a novel system architecture for LLVM-based modeling and simulation of custom hardware accelerators integrated into the gem5 framework. It overcomes the inherent limitations of state-of-the-art trace-based pre-register-transfer level (RTL) simulators by offering a truly "execute-in-execute" LLVM-based model. It enables scalable modeling of multiple dynamically interacting accelerators with full-system simulation support. To create long-term sustainable expansion compatible with the gem5 system framework, gem5-SALAM offers a general-purpose and modular communication interface and memory hierarchy integrated into the gem5 ecosystem, streamlining designing and modeling accelerators for new and emerging applications. gem5-SALAMv2 expands upon the framework established in gem5-SALAMv1 with improved LLVM-based elaboration and simulation, improved and more extensible system integration, and new automations to simplify rapid prototyping and design space exploration. 1 1 Conference Paper Extension: This work extends the work presented in gem5-SALAM: A System Architecture for LLVM-based Accelerator Modeling from MICRO 2020 (Rogers et al., 2020). This work expands on the aforementioned work by revamping the gem5-SALAM internals to provide more robust and extensible simulations, introducing new automation tools for expanding and simplifying design space exploration, and demonstrates the new capabilities of gem5-SALAMv2 by exploring multiple configurations of simple neural network architectures. Validation on the MachSuite (Reagen et al., 2014) benchmarks presents a timing estimation error of less than 1% against the Vivado High-Level Synthesis (HLS) tool. Results also show less than a 4% area and power estimation error against Synopsys Design Compiler. Additionally, system validation against implementations on an Ultrascale+ ZCU102 shows an average end-to-end timing error of less than 2%. Lastly, we demonstrate the upgraded capabilities of gem5-SALAMv2 by exploring accelerator platforms for two deep neural networks, LeNet5 and MobileNetv2. In these explorations, we demonstrate how gem5-SALAMv2 can simulate such systems and guide architectural optimizations for these types of accelerator-rich architectures. 2 2 The most up-to-date version of gem5-SALAMv2 is available at https://github.com/TeCSAR-UNCC/gem5-SALAM.. [ABSTRACT FROM AUTHOR]

Generative artificial intelligence (GAI) is a rapidly growing field with a wide range of applications. In this paper, a thorough examination of the research landscape in GAI is presented, encompassing a comprehensive overview of the prevailing themes and topics within the field. The study analyzes a corpus of 1319 records from Scopus spanning from 1985 to 2023 and comprises journal articles, books, book chapters, conference papers, and selected working papers. The analysis revealed seven distinct clusters of topics in GAI research: image processing and content analysis, content generation, emerging use cases, engineering, cognitive inference and planning, data privacy and security, and Generative Pre-Trained Transformer (GPT) academic applications. The paper discusses the findings of the analysis and identifies some of the key challenges and opportunities in GAI research. The paper concludes by calling for further research in GAI, particularly in the areas of explainability, robustness, cross-modal and multi-modal generation, and interactive co-creation. The paper also highlights the importance of addressing the challenges of data privacy and security in GAI and responsible use of GAI. [ABSTRACT FROM AUTHOR]

Digital platform ecosystem governance refers to a platform owner's decisions and mechanisms that seek to influence complementors and users to build and sustain an ecosystem. The relevance of digital platform ecosystem governance is broadly acknowledged by researchers. However, the extant body of research is fragmented, and varied terminologies are employed, leading to challenges in identifying and recognizing results across different fields. This article provides a multidisciplinary and systematic literature review with the aim of consolidating knowledge on this important topic. Based on an analysis of 103 journal articles and conference papers, this review synthesizes the literature into a conceptual model with five building blocks of platform ecosystem governance. The model aims to create a robust foundation for researchers approaching the topic for the first time and conducting subsequent research. The conceptual model also offers practical guidance for governing ecosystems in a structured manner. Finally, this article provides a research agenda with five areas for future investigation. • Multidisciplinary, systematic literature review on platform ecosystem governance. • Synthesis of the literature into a conceptual model with five building blocks. • Robust foundation for understanding the topic and for future research. • Research agenda with five areas and questions for future investigation. [ABSTRACT FROM AUTHOR]

In applications of AI systems where exact definitions of the important notions of the application domain are hard to come by, the use of traditional logic-based knowledge representation languages such as Description Logics may lead to very large and unintuitive definitions, and high complexity of reasoning. To overcome this problem, we define new concept constructors that allow us to define concepts in an approximate way. To be more precise, we present a family τ EL (m) of extensions of the lightweight Description Logic EL that use threshold constructors for this purpose. To define the semantics of these constructors we employ graded membership functions m , which for each individual in an interpretation and concept yield a number in the interval [ 0 , 1 ] expressing the degree to which the individual belongs to the concept in the interpretation. Threshold concepts C ⋈ t for ⋈ ∈ { < , ≤ , > , ≥ } then collect all the individuals that belong to C with degree ⋈ t. The logic τ EL (m) extends EL with threshold concepts whose semantics is defined relative to a function m. To construct appropriate graded membership functions, we show how concept measures ∼ (which are graded generalizations of subsumption or equivalence between concepts) can be used to define graded membership functions m ∼. Then we introduce a large class of concept measures, called simi-d , for which the logics τ EL (m ∼) have good algorithmic properties. Basically, we show that reasoning in τ EL (m ∼) is NP/coNP-complete without TBox, PSpace-complete w.r.t. acyclic TBoxes, and ExpTime-complete w.r.t. general TBoxes. The exception is the instance problem, which is already PSpace-complete without TBox w.r.t. combined complexity. While the upper bounds hold for all elements of simi-d , we could prove some of the hardness results only for a subclass of simi-d. This article considerably improves on and generalizes results we have shown in three previous conference papers and it provides detailed proofs of all our results. [ABSTRACT FROM AUTHOR]