Platform: Apple
  Device: AMD Radeon HD - FirePro D700 Compute Engine
    Driver version  : 1.2 (Mar 16 2017 18:19:56) (Macintosh)
    Compute units   : 32
    Clock frequency : 850 MHz

    Global memory bandwidth (GBPS)
      float   : 182.39
      float2  : 189.53
      float4  : 195.73
      float8  : 101.98
      float16 : 53.33

    Single-precision compute (GFLOPS)
      float   : 2593.97
      float2  : 2591.22
      float4  : 2584.96
      float8  : 2572.53
      float16 : 2543.34

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 654.65
      double2  : 654.62
      double4  : 653.86
      double8  : 652.82
      double16 : 650.63

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 11.12
      enqueueReadBuffer          : 11.92
      enqueueMapBuffer(for read) : 94.12
        memcpy from mapped ptr   : 6.79
      enqueueUnmap(after write)  : 7550.93
        memcpy to mapped ptr     : 7.64

    Kernel launch latency : 9.63 us

  Device: AMD Radeon HD - FirePro D700 Compute Engine
    Driver version  : 1.2 (Mar 16 2017 18:19:56) (Macintosh)
    Compute units   : 32
    Clock frequency : 850 MHz

    Global memory bandwidth (GBPS)
      float   : 184.77
      float2  : 191.74
      float4  : 197.64
      float8  : 102.48
      float16 : 53.56

    Single-precision compute (GFLOPS)
      float   : 2599.49
      float2  : 2594.67
      float4  : 2590.64
      float8  : 2576.32
      float16 : 2547.82

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 654.91
      double2  : 654.89
      double4  : 654.75
      double8  : 654.34
      double16 : 651.34

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 10.05
      enqueueReadBuffer          : 9.31
      enqueueMapBuffer(for read) : 85.90
        memcpy from mapped ptr   : 6.73
      enqueueUnmap(after write)  : 7389.83
        memcpy to mapped ptr     : 7.70

    Kernel launch latency : 10.21 us