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 30 for “"memory bottleneck"”.

  1. Addressing the memory bottleneck in packet processing systems

    … processing aggravates the well-known problem of memory bottleneck. To overcome the memory bottleneck, today’s packet processing systems support a wide-range of mechanisms such as exposed memory hierarchy, multithreading, and non-blocking multi-word accesses. However, supporting many such …

    texas Repository record for Addressing the memory bottleneck in packet processing systems (opens in a new tab)

  2. Mitigating the Memory Bottleneck with Machine Learning-Driven and Data-Aware Microarchitectural Techniques

    … the storage and retrieval capabilities of modern memory systems, making memory the dominant performance and energy-efficiency bottleneck in computing systems. Although numerous microarchitectural techniques attempt to hide or tolerate long memory access latency, rapidly growing data footprints …

    ethz Repository record for Mitigating the Memory Bottleneck with Machine Learning-Driven and Data-Aware Microarchitectural Techniques (opens in a new tab)

  3. ReMoDeL-FPGA: Reconfigurable Memory-centric Array Processor Architecture for Deep-Learning Acceleration on FPGA

    … data-dependent, requiring frequent access to memory. As a result, these applications suffer the most from the memory bottleneck of the von Neumann architectures. There is an imminent need for memory-centric architectures for deep-learning and big-data analytic applications that are memory …

    arkansas Repository record for ReMoDeL-FPGA: Reconfigurable Memory-centric Array Processor Architecture for Deep-Learning Acceleration on FPGA (opens in a new tab)

  4. Performance Modeling and Enhancement of the Structured Memory Access Architecture

    The structured memory access (SMA) architecture solves the traditional von Neumann processor-memory bottleneck by utilizing the structure of a program and the regular ways in which it accesses structured data in memory. The SMA machine consists of two pipelined processors: the computation processor …

    uiuc Repository record for Performance Modeling and Enhancement of the Structured Memory Access Architecture (opens in a new tab)

  5. Convergence Analysis of Segmental Refinement Multigrid

    … of unknowns on refined domains can create a memory bottleneck. The memory requirements can also reduce the efficiency of the algorithm due to the communication required between cores. I analyze Segmental Refinement Multigrid (SRMG) a modification to multigrid methods that reduces …

    rice Repository record for Convergence Analysis of Segmental Refinement Multigrid (opens in a new tab)

  6. Performance Characteristics of an Intelligent Memory System

    The memory system is increasingly becoming a performance bottleneck. Several intelligent memory systems, such as the ActivePages, DIVA, and IRAM architectures, have been proposed to alleviate the processor-memory bottleneck. This thesis presents the Memory Arithmetic Unit and Interface (MAUI) …

    maryland Repository record for Performance Characteristics of an Intelligent Memory System (opens in a new tab)

  7. Dynamic KV cache management for efficient LLM inference

    … LLMs generate text, making them powerful but memory-intensive. To accelerate generation, LLMs cache previously computed information in a Key-Value (KV) cache. As task complexity and sequence lengths grow, the KV cache becomes the primary memory bottleneck for LLM inference, limiting their …

    texas Repository record for Dynamic KV cache management for efficient LLM inference (opens in a new tab)

  8. IMPROVING PERFORMANCE OF MODERN MEMORY SUBSYSTEM

    … computing systems are largely dominated by its memory system. The memory system's performance has become critical as with the scaling of the device, the access latency of physical memory i.e. Dynamic random access memory (DRAM) has not scaled well. In the last two decades the capacity of DRAM …

    nus Repository record for IMPROVING PERFORMANCE OF MODERN MEMORY SUBSYSTEM (opens in a new tab) Download file for IMPROVING PERFORMANCE OF MODERN MEMORY SUBSYSTEM (opens in a new tab)

  9. Dynamically Reconfigurable Active Cache Modeling

    … Employing cache, particularly L1, can speed up memory accesses, reduce the effects of memory bottleneck and consequently improve the system performance; however, efficient design of a cache for embedded systems requires fast and early performance modeling. Our proposed model is cycle accurate …

    concordia Download PDF for Dynamically Reconfigurable Active Cache Modeling (opens in a new tab)

  10. Improving Memory Performance for Both High Performance Computing and Embedded/Edge Computing Systems

    CPU-memory bottleneck is a widely recognized problem. It is known that majority of high performance computing (HPC) database systems are configured with large memories and dedicated to process specific workloads like weather prediction, molecular dynamic simulations etc. My research on optimal …

    unt Repository record for Improving Memory Performance for Both High Performance Computing and Embedded/Edge Computing Systems (opens in a new tab)

  11. A study of thread-local garbage collection for multi-core systems

    … performance. One cause of poor scalability is memory bandwidth becoming saturated as processors contend with each other for memory bus use. More multi-core systems have a non-uniform memory architecture and placement of threads and the data they use is important in tackling this problem. …

    kent Download PDF for A study of thread-local garbage collection for multi-core systems (opens in a new tab)

  12. Practical irregular prefetching

    Memory accesses continue to be a performance bottleneck for many programs, and prefetching is an effective and widely used method for alleviating the memory bottleneck. However, prefetching can be difficult for irregular workloads, which the hardware has no clear patterns like sequential or strided …

    texas Repository record for Practical irregular prefetching (opens in a new tab)

  13. PBWT-based Methods for Biobank-Scale Haplotype Data Analysis

    … haplotype matches. Furthermore, we address the memory bottleneck of dynamic PBWT algorithms and develop a dynamic compressed PBWT called Dynamic μ-PBWT. We run-length compress the PBWT columns and store them in B+ trees for efficient updates. This enables efficient insertion/deletion of a …

    ucf Repository record for PBWT-based Methods for Biobank-Scale Haplotype Data Analysis (opens in a new tab)

  14. Fast and exact geodesic computation using Edge-based Windows Grouping.

    … it is still challenging to deal with the speed, memory cost and accuracy of the geodesic computation at the same time. This thesis addresses the aforementioned challenge by proposing the Edge- based Windows Grouping (EWG) technique. With the local geodesic information encoded in a “window”, EWG …

    bournemouth Download PDF for Fast and exact geodesic computation using Edge-based Windows Grouping. (opens in a new tab)

  15. Active management of Cache resources

    … The first set of challenges relates to the memory bottleneck. As the focus shifts from scaling processor frequency to scaling the number of cores, performance growth demands increasing die area. Scaling the number of cores also places a concurrent area demand in the form of larger caches. …

    gatech Repository record for Active management of Cache resources (opens in a new tab)

  16. Exploiting language abstraction to optimize memory efficiency

    … ease of development. These languages abstract memory management from the programmer, which can introduce time and space overhead but also provide opportunities for dynamic optimization. Optimizing memory performance is in part paramount because hardware is reaching physical limits. Recent …

    texas Repository record for Exploiting language abstraction to optimize memory efficiency (opens in a new tab)

  17. Determining Visual Motion in the Deep Learning Era

    … are also limited by the large consumption of GPU memory. An intermediate representation named cost volume is often employed which scales quadratically with the number of pixels. This 4D representation acts as a memory bottleneck for modern optical flow approaches, which prevents scaling up to …

    anu Repository record for Determining Visual Motion in the Deep Learning Era (opens in a new tab)

  18. Enabling large-scale seismic data acquisition, processing and waveform-inversion via rank-minimization

    … sensing to mitigate the computational and memory bottleneck of seismic processing workflows such as missing-trace interpolation, source separation and wave-equation based inversion for large-scale 3- and 5-D seismic data. For interpolation and source separation using rank-minimization, I …

    ubc Repository record for Enabling large-scale seismic data acquisition, processing and waveform-inversion via rank-minimization (opens in a new tab) Download PDF for Enabling large-scale seismic data acquisition, processing and waveform-inversion via rank-minimization (opens in a new tab)

  19. Machine learning applications for porous media

    … a limited data is available, how to overcome the memory bottleneck of hardware, and how to bridge information from multiple scales. The goal of the models proposed in here is not to replace simulations or laboratory experiments, but to complement them. Throughout this dissertation, it is shown how …

    texas Repository record for Machine learning applications for porous media (opens in a new tab)

  20. Νευρομορφικές υπολογιστικές τεχνικές σε εφαρμογές τηλεπικοινωνιών και επεξεργασίας εικόνας

    … data. Key challenges are the interconnection of memory and processing over slow electrical channels and the large number of electro-optical conversions at the interconnects. In recent years, the distance between information transfer and information processing has been significantly reduced with …

    greece Repository record for Νευρομορφικές υπολογιστικές τεχνικές σε εφαρμογές τηλεπικοινωνιών και επεξεργασίας εικόνας (opens in a new tab)

Page 1 of 2