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 12 of 12 for “"SIMD architectures"”.

  1. Efficient automatic code generation focused on deep learning for SIMD architectures

    [ES] La Multiplicación de matrices (o en inglés General Matrix Multiplication (GEMM)) es un componente fundamental en la computación científica y en los marcos actuales dentro del campo del Deep Learning (DL). Las implementaciones actuales de la GEMM están escritas en su mayoría en C, aunque su …

    upv Repository record for Efficient automatic code generation focused on deep learning for SIMD architectures (opens in a new tab)

  2. Throughput Optimizations for Irregular Dataflow Streaming Applications on Wide-SIMD Architectures

    … candidates for parallel acceleration using wide-SIMD vector processors, such as graphics cards. However, many of these streaming applications exhibit irregular dataflow, where the number of outputs per input from any computation within the application is data dependent and unknown a priori. To …

    wustl Repository record for Throughput Optimizations for Irregular Dataflow Streaming Applications on Wide-SIMD Architectures (opens in a new tab)

  3. Efficiently and Transparently Maintaining High SIMD Occupancy in the Presence of Wavefront Irregularity

    … detection, and big graph algorithms. With wide SIMD, lightweight threads, and low-cost thread-context switching, wide-SIMD architectures such as GPUs allow considerable flexibility in the way application work is assigned to threads. However, irregular applications are challenging to map …

    wustl Repository record for Efficiently and Transparently Maintaining High SIMD Occupancy in the Presence of Wavefront Irregularity (opens in a new tab)

  4. Compiling Data Dependent Control Flow on SIMD GPUs

    … units. Often these units are implemented as SIMD processors employing parallel pipelines. Data dependent conditional execution on SIMD architectures implemented using processor idling is inefficient. I propose a multi-pass approach based on conditional streams which allows dynamic load …

    waterloo Repository record for Compiling Data Dependent Control Flow on SIMD GPUs (opens in a new tab)

  5. Accelerating Tree-Based Irregular Applications by Exploiting SIMD Parallelism

    … of data level parallelism (DLP), and modern SIMD architectures, such as GPUs and the SIMD instruction extension to CPUs, appear well-suited. However, due to the high irregularity in control flow and data accesses of these applications, exploiting DLP on SIMD architectures is difficult and …

    unsw Repository record for Accelerating Tree-Based Irregular Applications by Exploiting SIMD Parallelism (opens in a new tab)

  6. Vector coprocessor sharing techniques for multicores: performance and energy gains

    … science many years ago. All commercial computing architectures on the market today contain some form of vector or SIMD processing. Many high-performance and embedded applications, often dealing with streams of data, cannot efficiently utilize dedicated vector processors for various reasons: …

    njit Repository record for Vector coprocessor sharing techniques for multicores: performance and energy gains (opens in a new tab)

  7. New motion estimation techniques and their SIMD implementations for video coding

    … to utilize the single instruction multiple data (SIMD) technique to accelerate a block motion estimation process, and to develop a new fast block motion estimation algorithm suitable for implementation using SIMD architecture. A method to detect blocks that are stationary between successive …

    concordia Download PDF for New motion estimation techniques and their SIMD implementations for video coding (opens in a new tab)

  8. Storage Management for Embedded SIMD Processors

    SIMD parallelism offers a high performance and efficient execution approach for today's broad range of portable multimedia consumer products. However, new methods are needed to meet the complex demands of high performance, embedded systems. This research explores new storage management techniques …

    gatech Repository record for Storage Management for Embedded SIMD Processors (opens in a new tab)