Citation
William Plishker. "Functional DIF". Talk or presentation, 5, June, 2008.

Abstract
Dataflow formalisms have provided designers of digital signal processing systems with analysis and optimizations for many years. As system complexity increases, designers are relying on more types of dataflow models to describe applications while retaining these implementation benefits. The semantic range of DSP-oriented dataflow models has expanded to cover heterogeneous models and dynamic applications, but efficient design, simulation, and scheduling of such applications has not. To facilitate implementing heterogeneous applications, we present a new dataflow model of computation that supports dynamic and static dataflow-based heterogeneous system design. With this semantic foundation, we show how actors designed in other dataflow models are directly supported by the functional dataflow interchange format (DIF) package, allowing system designers to immediately compose and simulate actors from different models. Schedules can then be expressed using a standard representation (the generalized schedule tree) which accommodates these dataflow models. Using an example, we show how this approach can be applied to quickly describe and functionally simulate a heterogeneous dataflow-based application.

Electronic downloads

• DIF%20talk.pdf · application/pdf · 555 kbytes
• DIF%20talk%20-%20chess.ppt · application/vnd.ms-powerpoint · 1242 kbytes

• HTML

  William Plishker. <a
  href="http://chess.eecs.berkeley.edu/pubs/431.html"
  ><i>Functional DIF</i></a>, Talk or
  presentation,  5, June, 2008.

• Plain text

  William Plishker. "Functional DIF". Talk or
  presentation,  5, June, 2008.

• BibTeX

  @presentation{Plishker08_FunctionalDIF,
      author = {William Plishker},
      title = {Functional DIF},
      day = {5},
      month = {June},
      year = {2008},
      abstract = {Dataflow formalisms have provided designers of
                digital signal processing systems with analysis
                and optimizations for many years. As system
                complexity increases, designers are relying on
                more types of dataflow models to describe
                applications while retaining these implementation
                benefits. The semantic range of DSP-oriented
                dataflow models has expanded to cover
                heterogeneous models and dynamic applications, but
                efficient design, simulation, and scheduling of
                such applications has not. To facilitate
                implementing heterogeneous applications, we
                present a new dataflow model of computation that
                supports dynamic and static dataflow-based
                heterogeneous system design. With this semantic
                foundation, we show how actors designed in other
                dataflow models are directly supported by the
                functional dataflow interchange format (DIF)
                package, allowing system designers to immediately
                compose and simulate actors from different models.
                Schedules can then be expressed using a standard
                representation (the generalized schedule tree)
                which accommodates these dataflow models. Using an
                example, we show how this approach can be applied
                to quickly describe and functionally simulate a
                heterogeneous dataflow-based application. },
      URL = {http://chess.eecs.berkeley.edu/pubs/431.html}
  }
  

Posted by Hiren Patel on 6 Jun 2008.
Groups: chess
For additional information, see the Publications FAQ or contact webmaster at chess eecs berkeley edu.