Platform: AMD Accelerated Parallel Processing
  Device: Tonga
    Driver version  : 1912.5 (VM) (Linux x64)
    Compute units   : 32
    Clock frequency : 1040 MHz

    Global memory bandwidth (GBPS)
      float   : 159.85
      float2  : 168.05
      float4  : 168.71
      float8  : 88.07
      float16 : 44.60

    Single-precision compute (GFLOPS)
      float   : 4178.26
      float2  : 4174.39
      float4  : 4164.47
      float8  : 4142.61
      float16 : 4095.45

    Double-precision compute (GFLOPS)
      double   : 263.50
      double2  : 263.37
      double4  : 263.09
      double8  : 262.54
      double16 : 261.39

    Integer compute (GIOPS)
      int   : 842.52
      int2  : 842.43
      int4  : 842.27
      int8  : 841.95
      int16 : 841.28

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 0.18
      enqueueReadBuffer          : 0.21
      enqueueMapBuffer(for read) : 4.22
        memcpy from mapped ptr   : 1.12
      enqueueUnmap(after write)  : 0.18
        memcpy to mapped ptr     : 1.11

    Kernel launch latency : 63.50 us