Platform: AMD Accelerated Parallel Processing
  Device: Spectre
    Driver version  : 1573.4 (VM) (Linux x64)
    Compute units   : 8
    Clock frequency : 720 MHz

    Global memory bandwidth (GBPS)
      float   : 27.21
      float2  : 28.08
      float4  : 22.67
      float8  : 15.10
      float16 : 8.28

    Single-precision compute (GFLOPS)
      float   : 733.13
      float2  : 732.25
      float4  : 730.70
      float8  : 726.86
      float16 : 718.54

    Double-precision compute (GFLOPS)
      double   : 45.98
      double2  : 45.95
      double4  : 45.91
      double8  : 45.81
      double16 : 45.61

    Integer compute (GIOPS)
      int   : 146.82
      int2  : 146.91
      int4  : 146.03
      int8  : 145.64
      int16 : 146.70

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 8.42
      enqueueReadBuffer          : 5.40
      enqueueMapBuffer(for read) : 4357.72
        memcpy from mapped ptr   : 3.76
      enqueueUnmap(after write)  : 8918.12
        memcpy to mapped ptr     : 3.54

    Kernel launch latency : 33.79 us

  Device: AMD A10-7850K APU with Radeon(TM) R7 Graphics
    Driver version  : 1573.4 (sse2,avx,fma4) (Linux x64)
    Compute units   : 4
    Clock frequency : 4000 MHz

    Global memory bandwidth (GBPS)
      float   : 9.72
      float2  : 10.97
      float4  : 12.09
      float8  : 11.17
      float16 : 14.40

    Single-precision compute (GFLOPS)
      float   : 3.12
      float2  : 5.69
      float4  : 12.62
      float8  : 22.69
      float16 : 44.04

    Double-precision compute (GFLOPS)
      double   : 3.11
      double2  : 6.28
      double4  : 11.28
      double8  : 23.00
      double16 : 28.83

    Integer compute (GIOPS)
      int   : 6.16
      int2  : 4.10
      int4  : 13.76
      int8  : 26.61
      int16 : 30.75

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 8.31
      enqueueReadBuffer          : 5.45
      enqueueMapBuffer(for read) : 10324.44
        memcpy from mapped ptr   : 3.76
      enqueueUnmap(after write)  : 19311.90
        memcpy to mapped ptr     : 3.56

    Kernel launch latency : 11.89 us