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 3 of 3 for “"block floating point"”.

  1. Complex block floating-point format with box encoding in communication systems

    … in communication systems design. A complex block floating-point format with box encoding is proposed to encode an array of complex numbers that has better numeric resolution than its IEEE-754 counterpart when the same number of bits are allocated to the dominant value in the array. It is …

    texas Repository record for Complex block floating-point format with box encoding in communication systems (opens in a new tab)

  2. Accelerating convolutional neural networks through enhanced designs

    … neural networks use the standard IEEE 32-bit floating point format. However, due to the significant redundancy of these models, often only a fourth of the bitwidth is necessary, and in many cases this can be pushed down to even 2 or 1 bit. We present three papers exploring this idea. Our first …

    oxford Repository record for Accelerating convolutional neural networks through enhanced designs (opens in a new tab)

  3. A Custom RTL Domain-Specific GEMM Accelerator on Cyclone V SoC: Dynamic Precision Selection and Tiling, from RTL Design to Linux Bring-Up for Edge AI Inference and Predictive Control

    … First, I assess runtime precision selection. A floating-point-capable processing element requires 12.8x the area and 7.5x the dynamic power of an integer element, although it uses half the hard-multiplier resource. Two independent tools locate the expense in the per-element floating-point …

    washington Repository record for A Custom RTL Domain-Specific GEMM Accelerator on Cyclone V SoC: Dynamic Precision Selection and Tiling, from RTL Design to Linux Bring-Up for Edge AI Inference and Predictive Control (opens in a new tab)