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 65 for “"Reduced-precision"”.

  1. Reconfigurable implementation with reduced precision of massive MIMO systems

    … coarsely quantized. Digital MMIMO receivers with reduced word-length for massive multiple-input multiple-output(MIMO) systems are proposed. The assessment of the bit error rate and extra power needed to compensate for the information loss due to the coarse quantization is assessed. The …

    uvic Repository record for Reconfigurable implementation with reduced precision of massive MIMO systems (opens in a new tab)

  2. Analytical guarantees for reduced precision fixed-point margin hyperplane classifiers

    … can effectively be traded off each other through precision. We analytically capture this trade-off by means of bounds on the precision requirements of general margin hyperplane classifiers. In addition, we propose a principled precision reduction scheme based on the trade-off between input and …

    uiuc Repository record for Analytical guarantees for reduced precision fixed-point margin hyperplane classifiers (opens in a new tab)

  3. Reduced-precision arithmetic in numerical weather prediction with an emphasis on data assimilation

    … In this thesis we explore the notion of reduced-precision arithmetic to accelerate Earth-System models, specifically those used in data assimilation and in numerical weather prediction.</p> <p>We first conduct data assimilation experiments with the Lorenz '96 toy atmospheric system, using …

    oxford Repository record for Reduced-precision arithmetic in numerical weather prediction with an emphasis on data assimilation (opens in a new tab)

  4. Use of the Reduced Precision Redundancy (RPR) method in a radix4 FFT implementation

    Reduced precision redundancy (RPR), as a new method for improving fault tolerance in FPGAs, appears promising in replacing triple modular redundancy (TMR) to prevent the single event effects due to radiation in arithmetic processes. As a test of this approach, the RPR technique was used to …

    nps Repository record for Use of the Reduced Precision Redundancy (RPR) method in a radix4 FFT implementation (opens in a new tab)

  5. Mitigating computational bottlenecks in deep neural networks through reduced-precision hardware and software

    … product of 130x (24x) compared with full-precision ADCs (sparse low-bit ADCs) while maintaining near-software accuracy in benchmark classification problems. To mitigate the accuracy loss characteristic of aggressive quantization, we offered compensatory training methods, including …

    iastate Repository record for Mitigating computational bottlenecks in deep neural networks through reduced-precision hardware and software (opens in a new tab)

  6. Reduced precision redundancy applied to arithmetic operations in field programmable gate arrays for satellite control and sensor systems

    … space vehicles and develops rules for applying Reduced Precision Redundancy (RPR) as a new method of fault tolerance in Field Programmable Gate Arrays against Single Event Effects due to radiation on orbit. RPR was discovered by Snodgrass in 2006 and was first demonstrated using the single-input …

    nps Repository record for Reduced precision redundancy applied to arithmetic operations in field programmable gate arrays for satellite control and sensor systems (opens in a new tab)

  7. Exploiting Data Characteristics in The Design of Accelerators for Deep Learning

    … this thesis examines the sparsity and precision requirements of Deep Convolutional Neural Networks, which comprise multiple layers of matrix-vector product computations. We propose a profiling method to find per layer reduced precision configurations while maintaining high …

    toronto-retro Repository record for Exploiting Data Characteristics in The Design of Accelerators for Deep Learning (opens in a new tab)

  8. Investigating the potential for improving the accuracy of weather and climate forecasts by varying numerical precision in computer models

    … computational costs. This 'scale-selective reduced precision' technique builds on previous work that shows that weather models can be run with almost all real numbers represented in 32 bit precision or lower without any impact on forecast accuracy, challenging the paradigm that 64 bits of …

    oxford Repository record for Investigating the potential for improving the accuracy of weather and climate forecasts by varying numerical precision in computer models (opens in a new tab)

  9. MODEL ADAPTATION FOR EDGE AI

    … structured sparsity pattern. Thirdly, while low-precision integer formats are common for reduced precision arithmetic, these formats are inadequate for emerging workloads like transformers. We explore minifloats, reduced-precision floating-point formats that significantly reduce the memory …

    nus Repository record for MODEL ADAPTATION FOR EDGE AI (opens in a new tab) Download file for MODEL ADAPTATION FOR EDGE AI (opens in a new tab)

  10. Differentiating Ethernet devices using Normal Link Pulse with efficient computation and the impacts on performance

    … are more readily available and inexpensive. Reduced precision at each decimation was also tested. The design space survey indicates that trimming the time domain NLP records is beneficial to a certain point, and tracking the changes or drift of the signal has a great benefit. The design space …

    iastate Repository record for Differentiating Ethernet devices using Normal Link Pulse with efficient computation and the impacts on performance (opens in a new tab)

  11. Harnessing resilience: biased voltage overscaling for probabilistic signal processing

    … error rates compared to an energy equivalent reduced-precision solution. When applied to H.264 video decoding, a BIVOS solution is able to achieve a 33.9% reduction in energy consumption while maintaining a peak-signal-to-noise ratio of 35.0dB (compared to 14.3dB for a comparable …

    gatech Repository record for Harnessing resilience: biased voltage overscaling for probabilistic signal processing (opens in a new tab)

  12. Fast and low-precision learning in GPU-accelerated spiking neural network

    … in complex problems and are hard to operate at reduced precision. This paper demonstrates a GPU-accelerated SNN architecture that uses stochasticity in the STDP coupled with higher frequency input spike trains. The simulation results demonstrate 2 to 3 times faster learning compared to …

    gatech Repository record for Fast and low-precision learning in GPU-accelerated spiking neural network (opens in a new tab)

  13. Early soft error reliability assessment of convolutional neural networks executing on resource-constrained IoT edge devices

    … devices, which typically rely on reduced memory footprint and low-power components (e.g., microcontrollers and processors). While performance im provement, customized, and reduced-precision implementations of ML algorithms have been studied extensively, their susceptibility to soft …

    brazil-ufrgs Repository record for Early soft error reliability assessment of convolutional neural networks executing on resource-constrained IoT edge devices (opens in a new tab)

  14. Efficient Hierarchical Design in Reservoir Computing

    … extraction across temporal scales, and (3) a reduced-precision arithmetic pipeline that leverages high-precision training followed by low-bit inference (e.g., 16-bit fixed-point) to minimize resource overhead while preserving predictive fidelity. The proposed design systematically reduces …

    mississippi Repository record for Efficient Hierarchical Design in Reservoir Computing (opens in a new tab)

  15. Precision Exploration and Underflow Mitigation for BAPS Digital Predistortion

    … for hardware implementation. However, when reduced precision floating-point arithmetic is used, these sequential multiplication chains create dynamic range challenges. This thesis investigates how the number of mantissa bits (t), the number of exponent bits (w), and the number of BAPS basis …

    chalmers Repository record for Precision Exploration and Underflow Mitigation for BAPS Digital Predistortion (opens in a new tab)

  16. Reducing dynamic power consumption in next generation DS-CDMA mobile communication receivers

    … algorithm level, we investigate the impact of reduced precision and arithmetic complexity on the performance of the DS-CDMA RAKE receiver. We then present architectures for implementing the reference and reduced complexity DS-CDMA RAKE receivers, and analyze these architectures with respect to …

    rice Repository record for Reducing dynamic power consumption in next generation DS-CDMA mobile communication receivers (opens in a new tab)

  17. Automated Floating-Point Precision Analysis

    … techniques for automated floating-point precision analysis. The contributions include a software framework that enables floating-point program analysis at the binary level, as well as specific techniques for cancellation detection, mixed-precision configuration, and reduced-precision …

    maryland Repository record for Automated Floating-Point Precision Analysis (opens in a new tab)

  18. Exploring Simplified Reservoir Computing Systems for Resource-constrained Edge AI Hardware

    … cyclic connectivity pattern that enables reduced memory footprint, faster execution, and more straightforward hardware mapping without significantly sacrificing computational expressiveness. To further enhance the trade-off between nonlinearity and memory capacity—a critical aspect for …

    mississippi Repository record for Exploring Simplified Reservoir Computing Systems for Resource-constrained Edge AI Hardware (opens in a new tab)

  19. On-Orbit FPGA SEU Mitigation and Measurement Experiments on the Cibola Flight Experiment Satellite

    … modular redundancy, duplication with compare, reduced precision redundancy, and SDRAM and FPGA block memory scrubbing. Regional SEU rates and the change in CFE's SEU rate over time show the measurable, expected effects of the South Atlantic Anomaly and the cycle of solar activity on CFE's SEU …

    byu Repository record for On-Orbit FPGA SEU Mitigation and Measurement Experiments on the Cibola Flight Experiment Satellite (opens in a new tab) Download PDF for On-Orbit FPGA SEU Mitigation and Measurement Experiments on the Cibola Flight Experiment Satellite (opens in a new tab)

  20. Error -Tolerant Digital Signal Processing

    … noise tolerance (ANT) technique referred to as reduced precision redundancy (RPR), which can be easily integrated to any DSP system. Combined with voltage overscaling (VOS), RPR achieves desired performance with considerable energy savings. We extend forward predictor-based ANT and present …

    uiuc Repository record for Error -Tolerant Digital Signal Processing (opens in a new tab)

Page 1 of 4