Platform: AMD Accelerated Parallel Processing
  Device: Cayman
    Driver version : 1348.4 (Linux x64)
    Compute units  : 24

    Global memory bandwidth (GBPS)
      float   : 154.77
      float2  : 141.49
      float4  : 82.39
      float8  : 70.60
      float16 : 39.35

    Single-precision compute (GFLOPS)
      float   : 674.44
      float2  : 1345.68
      float4  : 2601.47
      float8  : 2586.69
      float16 : 2573.38

    Double-precision compute (GFLOPS)
      double   : 671.24
      double2  : 671.59
      double4  : 670.93
      double8  : 669.51
      double16 : 666.50

    Integer compute (GIOPS)
      int   : 337.36
      int2  : 448.62
      int4  : 537.68
      int8  : 536.19
      int16 : 533.12

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 3.53
      enqueueReadBuffer          : 4.43
      enqueueMapBuffer(for read) : 152.89
        memcpy from mapped ptr   : 4.40
      enqueueUnmap(after write)  : 1781.26
        memcpy to mapped ptr     : 4.42

    Kernel launch latency : 44.22 us