Platform: AMD Accelerated Parallel Processing
  Device: Hawaii
    Driver version  : 1598.5 (VM) (Linux x64)
    Compute units   : 40
    Clock frequency : 824 MHz

    Global memory bandwidth (GBPS)
      float   : 270.25
      float2  : 276.32
      float4  : 276.61
      float8  : 155.61
      float16 : 78.30

    Single-precision compute (GFLOPS)
      float   : 4149.75
      float2  : 4145.72
      float4  : 4135.53
      float8  : 4114.04
      float16 : 4066.81

    Double-precision compute (GFLOPS)
      double   : 2080.65
      double2  : 2078.84
      double4  : 2073.76
      double8  : 2062.56
      double16 : 2039.19

    Integer compute (GIOPS)
      int   : 1389.48
      int2  : 1389.25
      int4  : 1388.45
      int8  : 1387.98
      int16 : 1386.09

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 3.82
      enqueueReadBuffer          : 3.51
      enqueueMapBuffer(for read) : 170435.20
        memcpy from mapped ptr   : 6.42
      enqueueUnmap(after write)  : 275318.41
        memcpy to mapped ptr     : 6.47

    Kernel launch latency : 31.72 us