• TESTED FOR THE CUSTOMER’S DESIRED END SOLUTION.
  • [FIG11] TI TCI6487.
  • Examining architectures, programming




    Download 0,76 Mb.
    bet8/14
    Sana23.07.2024
    Hajmi0,76 Mb.
    #268329
    1   ...   4   5   6   7   8   9   10   11   ...   14
    Bog'liq
    TrendsInMulticoreDSP.Gatherer

    [FIG10] The Agere SP2603.




    In processor affinity, the SMP OS scheduler is modified to allow programmers to assign a certain process to a specific core. All other processes are then assigned by the OS. SMP Linux has features to allow

    AT THE END, WHAT MATTERS MOST IS THE ACTUAL PERFORMANCE ACHIEVED WHEN THE CHIPS ARE


    TESTED FOR THE CUSTOMER’S DESIRED END SOLUTION.
    removed from the code. The major features are directives that specify that a well-struc- tured region of code should be executed by a team of threads, who share in the work. Such regions may be nested. Work

    such modifications. A number of programming languages fol- lowing this approach have appeared to extend or replace C to better allow programmers to express parallelism. These include OpenMP [25], MPI [26], X10 [27], MCAPI [28], GlobalArrays
    [29], and Uniform Parallel C [30]. In addition, functional lan- guages such as Erlang [31] and Haskell [32] as well as stream languages such as ACOTES [33] and StreamIT [34] have been introduced. Several of these languages have been ported to multicore DSPs. OpenMP is an example of that. It is a widely adopted shared-memory, parallel-programming interface pro- viding high-level programming constructs that enable the user to easily expose an application’s task and loop-level parallelism in an incremental fashion. Its range of applicability was signifi- cantly extended by the addition of explicit tasking features. The user specifies the parallelization strategy for a program at a high level by annotating the program code; the implementa- tion works out the detailed mapping of the computation to the machine. It is the user’s responsibility to perform any code modifications needed prior to the insertion of OpenMP con- structs. In particular, OpenMP requires that dependencies that might inhibit parallelization are detected and where possible,


    VCP2
    [FIG11] TI TCI6487.
    sharing directives are provided to effect a distribution of work among the participating threads [35].
    Virtualization partitions the software and hardware into a set of virtual machines (VMs) that are assigned to the cores using a VM manager (VMM). This allows multiple operating systems to run on single or multiple cores. Virtualization works as a level of abstraction between the OS and the hardware. VirtualLogix employs virtualization technology using its VLX for embedded systems [36]. VLX announced support for TI single and multicore DSPs. It allows TI’s real-time OS (DSP/BIOS) to run concurrently with Linux. Therefore, DSP/BIOS is left to run critical tasks while other applications run on Linux.



    Download 0,76 Mb.
    1   ...   4   5   6   7   8   9   10   11   ...   14




    Download 0,76 Mb.

    Bosh sahifa
    Aloqalar

        Bosh sahifa



    Examining architectures, programming

    Download 0,76 Mb.