Changeset 104 for anr/section-3.2.tex


Ignore:
Timestamp:
Feb 8, 2010, 9:17:37 AM (14 years ago)
Author:
coach
Message:

pc

File:
1 edited

Legend:

Unmodified
Added
Removed
  • anr/section-3.2.tex

    r99 r104  
    99\end{figure}
    1010\begin{description}
    11 \item[HPC setup] Here the user splits the application into 2 parts: the host application
    12 which remains on a PC and the SoC application which migrates into a SoC.
    13 The framework provides a simulation model which allows an evaluation of the partitioning.
    14 \item[SoC design] In this phase,
     11\item[HPC setup:] During this step, the user splits the application into 2 parts: the host application
     12which remains on a PC and the SoC application which is mapped on the FPGA.
     13The COACH framework provides a SystemC simulation model of the whole system (PC+communication+FPGA-SoC) which allows a performance evaluation of the partitioning.
     14\item[SoC design:] In this phase,
    1515the user can obtain simulators for the SoC at different abstraction levels by giving to the COACH framework a SoC description. 
    1616This description consists of a process network corresponding to the SoC application,
     
    2020ASIP (the process runs on a SoC processor enhanced with dedicated instructions),
    2121and hardware (the process runs into a coprocessor that is generated by HLS and plugged on the SoC bus).
    22 \item[Application compilation] Once the SoC description is validated, COACH generates automatically
     22\item[Application compilation:] Once the SoC description is validated, COACH generates automatically
    2323an FPGA bitstream containing the hardware platform with the SoC application software and
    2424an executable containing the host application. The user can launch the application by
     
    2828% l'avancee scientifique attendue. Preciser l'originalite et le caractere
    2929% ambitieux du projet.
     30%FIXME == {NON ceci n'est pas une contribution scientifique. A re-ecrire}
    3031The main scientific contribution of the project is to unify various synthesis techniques
    3132(same input and output formats) allowing the user to swap without engineering effort
    32 from one to another and even to chain them. for instance, it will be possible to run loop transformations before synthesis.
     33from one to another and even to chain them. For instance, it will be possible to run loop transformations before synthesis.
    3334Another advantage of this framework is to provide different abstraction levels from
    3435a single description.
     
    4142and technological barriers.
    4243\begin{itemize}
    43 \item The main problem in HPC is the communication between the PC and the SoC.
    44 This problem has 2 aspects. The first one is the run-time efficiency. The second is its engineering  cost, especially if one want to refine an implementation
    45 at several abstract levels.
    46 \item The COACH design flow has a top-down approach. In such a case,
    47 the required performance of a coprocessor (clock frequency, maximum cycles for
    48 a given computation, power consumption, etc) are imposed by the other system
    49 components. The challenge is to allow user to control accurately the synthesis
    50 process. For instance, the clock frequency must not be a result of the RTL synthesis
    51 but a strict synthesis constraint.
    5244\item HLS tools are sensitive to the style in which the algorithm is written.
    5345In addition, they are are not integrated into an architecture and system
     
    6961to be solved, mainly to do with the construction of FIFO communication
    7062channels and with memory optimization.
     63\item The COACH design flow has a top-down approach. In such a case,
     64the required performance of a coprocessor (clock frequency, maximum cycles for
     65a given computation, power consumption, etc) are imposed by the other system
     66components. The challenge is to allow user to control accurately the synthesis
     67process. For instance, the clock frequency must not be a result of the RTL synthesis
     68but a strict synthesis constraint.
     69\item The main problem in HPC is the communication between the PC and the SoC.
     70This problem has 2 aspects. The first one is the run-time efficiency. The second is
     71its engineering  cost, especially if one want to refine an implementation
     72at several abstract levels.
    7173
    7274\end{itemize}
Note: See TracChangeset for help on using the changeset viewer.