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 25 for “"High-bandwidth memory"”.

  1. HIGH THROUGHPUT OPENCL ACCELERATOR FOR SPARSE ALGEBRA

    With the emergence of FPGA boards equipped with High Bandwidth Memory (HBM2), these boards have become more attractive for implementing memory-intensive applications. High-Level Synthesis (HLS) compilers also have eased programming of FPGAs, allowing developers to program FPGAs in high-level …

    gatech Repository record for HIGH THROUGHPUT OPENCL ACCELERATOR FOR SPARSE ALGEBRA (opens in a new tab)

  2. A design integration framework to enable the rapid generation of large-scale accelerator designs

    … with the growing number of designs utilizing high-bandwidth memory (HBM) to accelerate traditionally memory-bound tasks. The inherent complexity of such large systems imposes significant design and optimization challenges. Traditional system integration tools were not designed for this level …

    texas Repository record for A design integration framework to enable the rapid generation of large-scale accelerator designs (opens in a new tab)

  3. IMPROVING THE PERFORMANCE AND ENERGY EFFICIENCY OF EMERGING MEMORY SYSTEMS

    Modern main memory is primarily built using dynamic random access memory (DRAM) chips. As DRAM chip scales to higher density, there are mainly three problems that impede DRAM scalability and performance improvement. First, DRAM refresh overhead grows from negligible to severe, which limits DRAM …

    vcu Repository record for IMPROVING THE PERFORMANCE AND ENERGY EFFICIENCY OF EMERGING MEMORY SYSTEMS (opens in a new tab)

  4. A Framework for Comprehensive and Customizable Memory-Centric Workloads

    … putting tremendous demands on the system's memory bandwidth, energy, and predictability. Because main memory is the main bottleneck in such computing systems, researchers have proposed a number of novel memory solutions. However, because memory-centric benchmarks have been slow to emerge, …

    mcmaster Repository record for A Framework for Comprehensive and Customizable Memory-Centric Workloads (opens in a new tab)

  5. Improving system performance by utilizing heterogeneous memories

    … cannot scale anymore. In order to continue the memory scaling, vendors are beginning to use new emerging memory technologies such as die-stacked DRAM. Although each type of memory technology has advantages and disadvantages, none has characteristics that are identical to conventional DRAM. For …

    texas Repository record for Improving system performance by utilizing heterogeneous memories (opens in a new tab)

  6. Parallel Architecture Exploration for Data-intensive Applications

    … scalable and low-power solution to the emerging high-volume, low locality problem. Instead of parting from an existent computer architecture or trying to adapt and accelerate an existing implementation, our approach is to extract the essential requirements of data-intensive applications to turn …

    toronto-retro Repository record for Parallel Architecture Exploration for Data-intensive Applications (opens in a new tab)

  7. Ring-Based Resonant Standing Wave Oscillators for 3D Clocking Applications

    … tech-nique that can distribute and generate a high frequency, low skew, low power, and stable clock signal. By using through-silicon-vias, this type of standing wave oscillator can be used to gener-ate the clocking scheme for 3D integrated circuits. In this thesis, we propose the use of such 3D …

    tamu Repository record for Ring-Based Resonant Standing Wave Oscillators for 3D Clocking Applications (opens in a new tab)

  8. Leveraging Many-Core Technology for Deterministic Neutral Particle Transport at Extreme Scale

    … to leverage the increased parallelism and high bandwidth memory technologies from many-core devices. The deterministic discrete ordin- ates transport equation is an important equation which models the movement and interaction of neutral particles, such as neutrons, through materials. The …

    bristol Repository record for Leveraging Many-Core Technology for Deterministic Neutral Particle Transport at Extreme Scale (opens in a new tab) Download PDF for Leveraging Many-Core Technology for Deterministic Neutral Particle Transport at Extreme Scale (opens in a new tab)

  9. Investigation of processing condition on the reliability and performance of future memory devices

    … the total semiconductor market is comprised of memory technologies. However, due to the limits in scalability related to Moore's Law, the need for additional memory capacity and bandwidth is increasing. Currently, dynamic random-access memory (DRAM) is in high-volume but is reaching its scaling …

    gatech Repository record for Investigation of processing condition on the reliability and performance of future memory devices (opens in a new tab)

  10. Automated Program Profiling and Analysis for Managing Heterogeneous Memory Systems

    <p>Many promising memory technologies, such as non-volatile, storage-class memories and high-bandwidth, on-chip RAMs, are beginning to emerge. Since each of these new technologies present tradeoffs distinct from conventional DRAMs, next-generation systems are likely to include multiple tiers of …

    ut-knoxville Repository record for Automated Program Profiling and Analysis for Managing Heterogeneous Memory Systems (opens in a new tab)

  11. Architectures and Frameworks for Near-Memory Computing using 3D-Stacked Memories

    … generate vast amounts of data and demand high computational performance. The advent of advanced Artificial Intelligence (AI) models, particularly transformer-based Large Language Models (LLMs), has intensified these demands by processing enormous datasets for tasks such as text …

    qucosa-diss

  12. Cache-Conscious Sparse Matrix Dense Matrix Multiplication on GPUs

    Over the past decade, high-performance deep learning has evolved into a critical research domain, driven by the demand for efficient models and high inference throughput. Deep learning architectures have shifted from stacked convolutional layers to transformer-based models, while pruning techniques …

    lsu-thes Repository record for Cache-Conscious Sparse Matrix Dense Matrix Multiplication on GPUs (opens in a new tab)

  13. Modern vector architectures for high-performance computing

    … central processing units (CPUs) for the high-performance segment have had to adopt new approaches in order to deliver increasing performance. Clock frequency has increased little, but the number of cores per chip has increased by several times in a single decade. Inside each core, single …

    bristol Repository record for Modern vector architectures for high-performance computing (opens in a new tab) Download PDF for Modern vector architectures for high-performance computing (opens in a new tab)

  14. Modeling, design, materials, processes and reliability of multi-layer redistribution wiring layers on glass substrates for next generation of high-performance computing applications

    There is a growing demand for high performance computing with miniaturization in many electronic systems such as servers for cloud computing, accurate weather prediction, smart mobile and wearable devices and autonomous driving cars. The development of 2.5D silicon interposers in 2010 for …

    gatech Repository record for Modeling, design, materials, processes and reliability of multi-layer redistribution wiring layers on glass substrates for next generation of high-performance computing applications (opens in a new tab)

  15. Analytical Modeling and Efficient Implementation for Specialized Hardware Acceleration of Sparse Applications

    … a versatile sparse accelerator equipped with High Bandwidth Memory (HBM) to efficiently handle datasets of varying dimensions and density. Vesper accelerates four key SLA kernels: sparse-matrix dense-vector multiplication (SpMV), sparse-matrix dense-matrix multiplication (SpMM), sparse-matrix …

    cornell Repository record for Analytical Modeling and Efficient Implementation for Specialized Hardware Acceleration of Sparse Applications (opens in a new tab) Embargoed until 2028-06-18

  16. Architecting heterogeneous memory systems with 3D die-stacked memory

    … research is to efficiently enable 3D die-stacked memory and heterogeneous memory systems. 3D die-stacking is an emerging technology that allows for large amounts of in-package high-bandwidth memory storage. Die-stacked memory has the potential to provide extraordinary performance and energy …

    gatech Repository record for Architecting heterogeneous memory systems with 3D die-stacked memory (opens in a new tab)

  17. Scalable Access-Pattern Aware I/O Acceleration and Multi-Tiered Data Management for HPC and AI Workloads

    … workloads presents significant challenges for high-performance computing~(HPC) systems. These workloads generate massive data volumes at unprecedented velocities, straining the capabilities of existing memory hierarchies, I/O subsystems, and scheduling mechanisms. This dissertation addresses …

    rit Repository record for Scalable Access-Pattern Aware I/O Acceleration and Multi-Tiered Data Management for HPC and AI Workloads (opens in a new tab)

  18. Compiler analysis and optimization of memory management In modern processors

    … such as CPUs and GPUs have continued to deliver higher performance with increasing microarchitecture innovations in recent years. However, it is non-trivial for application developers to achieve anywhere close to peak performance on these platforms. Memory is the most common system performance …

    gatech Repository record for Compiler analysis and optimization of memory management In modern processors (opens in a new tab)

  19. Architecting high-performance, efficient, and scalable heterogeneous memory systems with 3D-DRAM

    Advancements in packaging technology enable high-bandwidth 3D-DRAM that mitigates the memory bandwidth wall. With the integration of 3D-DRAM and high-capacity memory, heterogeneous memory systems are able to satisfy the high memory-bandwidth demand of processors. However, traditional management …

    gatech Repository record for Architecting high-performance, efficient, and scalable heterogeneous memory systems with 3D-DRAM (opens in a new tab)

  20. Architectural-Physical Co-Design of 3D CPUs with Micro-Fluidic Cooling

    … design space of heterogeneously-integrated high-bandwidth CPUs. Vertical integration promises to provide the CPU architectures of the future by integrating high performance processors with on-chip high-bandwidth memory systems and highly connected network-on-chip structures. Such techniques …

    maryland Repository record for Architectural-Physical Co-Design of 3D CPUs with Micro-Fluidic Cooling (opens in a new tab)

Page 1 of 2