Global ETD Search

Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.

Results

Showing 1 to 20 of 20 for “"Dataflows"”.

  1. Democratizing scalable cloud applications: transactional stateful functions on streaming dataflows

    Web applications power almost every aspect of our digitalized society, from entertainment to web shopping, vacation planning and booking, online games, communication, work, and social interaction. However, building scalable and consistent Web applications in modern cloud environments requires …

    greece Repository record for Democratizing scalable cloud applications: transactional stateful functions on streaming dataflows (opens in a new tab)

  2. Understanding the Design Space of Dataflows for Graph Neural Network Accelerators

    … have been proposed which use particular dataflows for each phase(intra-phase) and different communication strategies between the two phases(inter-phase). Prior works on GNN Accelerators propose different optimizations in hardware and software in order to efficiently run Graph Neural …

    gatech Repository record for Understanding the Design Space of Dataflows for Graph Neural Network Accelerators (opens in a new tab)

  3. Architectures and Data Orchestration for Tensor Algebra Applications with Low-reuse Operations

    … reuse. We first propose a taxonomy to classify dataflows for pipelining between two operations, and a cost-model OMEGA. We then propose PipeOrgan, which extends this to a larger pipeline depth and proposes a new class of spatial organization strategies for parallel pipeline dataflows, which …

    gatech Repository record for Architectures and Data Orchestration for Tensor Algebra Applications with Low-reuse Operations (opens in a new tab)

  4. Understanding and Enforcing Privacy Compliance in Smart Homes

    … detects 532 apps with sensitive external dataflows, among which 408 (76.7%) apps have undisclosed data leaks. Moreover, 63.4% of the dataflows that involve health and wellness data are inconsistent with the practices disclosed in theapps’ privacy policies. Next, we focus on the compliance …

    ku Repository record for Understanding and Enforcing Privacy Compliance in Smart Homes (opens in a new tab)

  5. DESIGN SPACE EXPLORATION TOWARDS ENERGY EFFICIENT COARSE GRAINED RECONFIGURABLE ARRAYS

    … co-design. It maps variable-precision dataflows to uniform lower bit forms, simplifying hardware while supporting precision variability. Finally, the thesis presents PACE, a low-power SoC that integrates a RISC-V processor with a CGRA accelerator optimized for edge applications.

    nus Repository record for DESIGN SPACE EXPLORATION TOWARDS ENERGY EFFICIENT COARSE GRAINED RECONFIGURABLE ARRAYS (opens in a new tab) Download file for DESIGN SPACE EXPLORATION TOWARDS ENERGY EFFICIENT COARSE GRAINED RECONFIGURABLE ARRAYS (opens in a new tab)

  6. MEMORY-CENTRIC ARCHITECTURES FOR AI INFERENCE ACCELERATION VIA HARDWARE–SOFTWARE CO-DESIGN

    … digital IMC architecture that supports diverse dataflows in convolutional neural networks while reducing on-chip data movement. For large language models, the thesis introduces a hybrid IMC/NMC processing-communication fabric that assigns static weights, dynamic activations, and reductions to …

    nus Repository record for MEMORY-CENTRIC ARCHITECTURES FOR AI INFERENCE ACCELERATION VIA HARDWARE–SOFTWARE CO-DESIGN (opens in a new tab) Download file for MEMORY-CENTRIC ARCHITECTURES FOR AI INFERENCE ACCELERATION VIA HARDWARE–SOFTWARE CO-DESIGN (opens in a new tab)

  7. Domain Specific Memory Management for Large Scale Data Analytics

    … bottlenecks in the implementations of dataflows typically present in large-scale data analytics applications. In particular, efficient use of main memory has emerged as a critical aspect of dataflow implementation, due to the proliferation of multi-core architectures, as well as the …

    jhu Repository record for Domain Specific Memory Management for Large Scale Data Analytics (opens in a new tab)

  8. Automated Testing for Data-Intensive Scalable Computing

    … enable developers to write applications as dataflows composed of user-defined functions operating over large and often unstructured datasets. However, testing such applications and the underlying frameworks remains a significant challenge. Conventional testing techniques struggle in this …

    vt Repository record for Automated Testing for Data-Intensive Scalable Computing (opens in a new tab)

  9. Hardware Acceleration of Machine Learning and Combinatorial Optimization Problems

    … ML workloads exhibit dense and parallelizable dataflows, while CO problems involve irregular and constraint-driven computations. To address these characteristics, this dissertation explores the hardware acceleration of ML and CO workloads using application-specific integrated circuits (ASICs). …

    umn Repository record for Hardware Acceleration of Machine Learning and Combinatorial Optimization Problems (opens in a new tab)

  10. Improving Scalability of Deep Learning Accelerators Using 3D IC

    … reconfigurable compute arrays and adaptable dataflows. However, this flexibility introduces substantial physical design overhead, particularly in wirelength, area, power, and critical path timing. To address these bottlenecks, we propose architectural and physical design methodologies that …

    gatech Repository record for Improving Scalability of Deep Learning Accelerators Using 3D IC (opens in a new tab)

  11. Elastic dataflow processing on the cloud

    … transformed into data processing flows (dataflows). A logical question that arises is whether elasticity affects dataflow execution and in which way. It seems reasonable that the execution is faster when more resources are used, however the monetary cost is higher. This gives rise to the …

    greece Repository record for Elastic dataflow processing on the cloud (opens in a new tab)

  12. Near-Memory Computing Architectures for Scalable Edge AI Applications

    … configurable convolutional neuralnetwork (CNN) dataflows and achieves nearly two orders of magnitude higher throughput while sharply reducing internaldata movement relative to processor–memory baselines. To further improve efficiency under strict energy constraints, afamily of approximate …

    lund Repository record for Near-Memory Computing Architectures for Scalable Edge AI Applications (opens in a new tab) Download PDF for Near-Memory Computing Architectures for Scalable Edge AI Applications (opens in a new tab)

  13. Building Efficient Tensor Accelerators for Sparse and Irregular Workloads

    … functional sparse accelerator of different dataflows, e.g. inner product vs output product SpGEMM. The impact of Hard TACO is that it allows realistic architectural exploration of homogeneous and heterogeneous accelerators.

    gatech Repository record for Building Efficient Tensor Accelerators for Sparse and Irregular Workloads (opens in a new tab)

  14. Cross-Layer Design of Highly Scalable and Energy-Efficient AI Accelerator Systems Using Photonic Integrated Circuits

    … and provide flexible support for different dataflows. The system-level enhancements involve the integration of stochastic computing with PIC-based accelerators to break the inherent trade-off between scalability and supported bit precision. Additionally, applying stochastic computing …

    uky-thes Repository record for Cross-Layer Design of Highly Scalable and Energy-Efficient AI Accelerator Systems Using Photonic Integrated Circuits (opens in a new tab)

  15. Circuit architecture co-design for machine learning and 3D deep learning

    … a dedicated Gaussian rasterizer and optimized dataflows, the proposed architecture significantly improves the performance and efficiency of emerging 3D scene representation and 3DDL workloads.

    texas Repository record for Circuit architecture co-design for machine learning and 3D deep learning (opens in a new tab)

  16. Analyzing and improving MAESTRO's analytical modeling

    Deep learning accelerators efficiently execute deep learning applications through customization. However, designing specialized hardware takes considerable time and engineering effort. Design space exploration (DSE) tools automate the design of specialized accelerators by automatically evaluating …

    texas Repository record for Analyzing and improving MAESTRO's analytical modeling (opens in a new tab)

  17. Architecture circuit co-design for machine learning and combinatorial optimization problems

    … mismatch often exists among algorithmic dataflows, hardware architectures, and underlying circuit designs. General-purpose processors and standard Domain-Specific Architectures (DSAs) frequently impose rigid dataflow limitations. As a result, they fail to map efficiently to the irregular …

    texas Repository record for Architecture circuit co-design for machine learning and combinatorial optimization problems (opens in a new tab)