Sökning: "Floating-Point Precision"

Visar resultat 1 - 5 av 8 avhandlingar innehållade orden Floating-Point Precision.

  1. 1. Verification Based Failure Detection for Real-Time Java and Floating-Point Computations

    Författare :Gabriele Paganelli; Chalmers University of Technology; []
    Nyckelord :NATURVETENSKAP; NATURAL SCIENCES; Debugging; Floating-Point; Formal Methods; Testing; Real-Time Java; Coverage; Satisfiability Modulo Theories; Certification; Formal Specification;

    Sammanfattning : To err is human, and machines help us avoiding errors. This thesis presents two ways, based on verification technology, to prevent failures.Safety critical Real-Time Java applications usually need a certification. Tests fulfilling specific coverage criteria can be part of such certification, like in the DO-178 standard. LÄS MER

  2. 2. An Automated and Controlled Numerical Precision Reduction Framework for GPUs

    Författare :Alexandra Angerd; Chalmers University of Technology; []
    Nyckelord :NATURVETENSKAP; TEKNIK OCH TEKNOLOGIER; NATURAL SCIENCES; ENGINEERING AND TECHNOLOGY; Microarchitecture; Floating-Point Precision; Approximate Computing; Register File; GPU;

    Sammanfattning : Reducing the precision of floating-point values is an effective approach to achieve higher performance as well as higher energy-efficiency. This is especially true for GPUs, since many of its common tasks are inherently insensitive to precision-reduction. LÄS MER

  3. 3. Approximation and Compression Techniques to Enhance Performance of Graphics Processing Units

    Författare :Alexandra Angerd; Chalmers University of Technology; []
    Nyckelord :NATURVETENSKAP; TEKNIK OCH TEKNOLOGIER; NATURAL SCIENCES; ENGINEERING AND TECHNOLOGY; GPU; Compression; Approximate Computing; Register File; Machine Learning; Floating-Point Precision; Microarchitecture;

    Sammanfattning : A key challenge in modern computing systems is to access data fast enough to fully utilize the computing elements in the chip. In Graphics Processing Units (GPUs), the performance is often constrained by register file size, memory bandwidth, and the capacity of the main memory. LÄS MER

  4. 4. Performance and Energy Efficient Building Blocks for Network-on-Chip Architectures

    Författare :Sriram R. Vangal; Atila Alvandpour; Viktor Öwall; Linköpings universitet; []
    Nyckelord :NATURAL SCIENCES; NATURVETENSKAP; NATURVETENSKAP; NATURAL SCIENCES; Network-on-Chip Architectures; floating-point units; tiled-architectures; crossbar routers; multi-processor interconnection; Computer engineering; Datorteknik;

    Sammanfattning : The ever shrinking size of the MOS transistors brings the promise of scalable Network-on-Chip (NoC) architectures containing hundreds of processing elements with on-chip communication, all integrated into a single die. Such a computational fabric will provide high levels of performance in an energy efficient manner. LÄS MER

  5. 5. Statistical Compression Cache Designs

    Författare :Angelos Arelakis; Chalmers University of Technology; []
    Nyckelord :NATURVETENSKAP; NATURAL SCIENCES; Cache; Floating-point compression; Data Compression; Huffman Coding; Statistical Compression; Value Locality; Memory; Hybrid Compression;

    Sammanfattning : On-chip caches are essential as they bridge the growing speed-gap between off-chip memory and processors. To this end, processing cores are sacrificed for more cache space in the chip's real estate, possibly affecting the cache access time and power dissipation. LÄS MER