Now showing 1 - 3 of 3
  • Publication
    Overview of Software Level Power Consumption Models and Power Saving Techniques for Embedded DSP Processors
    Unlike DSP compilation for high performance, research for low power optimisation has received little attention, although power dissipation is a critical issue for mobile devices. This paper presents an overview of power consumption models and power saving techniques for embedded DSP processors applications and evaluates their application to the Texas Instruments TMS320VC5510 Digital Signal Processor. Software level power consumption models introduced in the literatureare presented, along with their advantages and disadvantages. Several power saving techniques are presented, discussing their relevance for the VC5510 processor architecture. The significance of various instruction components with respect to consumption are considered in detail.
  • Publication
    Dynamic Current Modelling at the Instruction Level
    Estimation of processor current consumption is important for the design of low power systems. This paper proposes a novel method for estimating the dynamic current consumption of a processor. The method models dynamic current as the output of a linear system excited by a signal comprised of the total current due to each instruction. System identification is performed by cross-correlation of a pseudo-random stimulus with the measured current. The method was applied to the Texas Instruments TMS320VC5510 DSP and was found to provide an average correlation of 93% between estimated and measured dynamic current across a range of benchmarks.
  • Publication
    Effect of Compiler Optimizations on DSP Processor Power and Energy Consumption
    This paper examines the effect of compiler optimizations on the energy usage and power consumption of DSP processor, specifically to the Texas Instruments TMS320VC5510. The effects of different levels of general and specific optimization on the energy and power consumption are measured for this processor. Given the special characteristics of DSP programs, the benchmark routines were selected from DSPStone, and some typical DSP applications. Finally, Texas Instruments library routines are compared with the compiled versions. The paper provides an analysis of the results together with recommendations for improving performance. The binaries used in this study were generated using the Texas Instrument C/C++ Compiler, which allows control over the whole set of optimizations.