Department of Electrical Engineering & Computer Science


ECM 6308: Current Topics in Parallel Processing

  • Catalog Description:
  • 3(3,0). PR: C.I. Research topics in parallel architectures, including, but not limited to, systolic architectures, wavefront arrays, interconnection networks, reconfigurable architectures and fast algorithms. May be used in the degree program a maximum of 2 times. Even Fall.

  • Spring 2015:
  • Topic: Post-Moore Architectures: Adaptive, Resilient, and Low Energy Designs
    Project:
    - Evolutionary Design of Approximate Circuits
    - Complimentary Memory Technologies to Reform the Homogeneity, Energy Landscape, and Volatility of Future Memory Hierarchies
    - A Survey of Imprecise Signal Processing
    - Power Wall and Dark Silicon: Multicore Solutions
    - Analyzing the Issues of Soft Error Rates(SER)in Technology Trend & Reliability for CMOS designs

  • Spring 2014-previous:
  • Topic: Post-Moore Architectures: Adaptive, Resilient, and Low Energy Designs

    Project:
    - N-Modular Redundancy Techniques for Fault Tolerance in Reconfigurable Logic Devices
    - Discussion of nMR-based Approaches for Fault-handing in SRAM-based FPGA Devices
    - Performance Analysis of Current Fault Management Techniques in FPGAs
    - An Overview of Different TMR Design Techniques with Comparison to Alternative Fault-Tolerant Design Methods for FPGAs
    - A Brief Survey of Techniques for Fault Tolerance in Field Programmable Devices
    - Fault Demotion Using Reconfigurable Slacks (FADRES)
    - Design-for-Diversity Using Case-Based and Inverted-Output Synthesis in Redundant Systems
    - Autonomous Built-in Self-Test Methods for SRAM Based FPGAs
    - Design and Adaptations to Mitigate Aging and Improve Reliability
    - Evolutionary Design of Approximate Circuits
    - Hybrid Discrete-Continuous Signal Processing: Employing Field-Programmable Analog Components for Energy-Sparing Computation
    - Variability-Aware Pruning of Approximate CMOS Logic
    - Tolerating Technology Scaling Challenges using Stochastic Computation
    - Timing Management Strategies: Comparing the use of Recovery, Masking, and Borrowing Strategies
    - Mitigating Variability from Technology Scaling


    University of Central Florida