Platform: Portable Computing Language
  Device: pthread-AMD EPYC 7662 64-Core Processor
    Driver version  : 1.8 (Linux x64)
    Compute units   : 128
    Clock frequency : 2597 MHz
64 warnings generated.

    Global memory bandwidth (GBPS)
      float   : 31.03
      float2  : 30.88
      float4  : 32.47
      float8  : 28.72
      float16 : 30.20

    Single-precision compute (GFLOPS)
      float   : 82.55
      float2  : 162.69
      float4  : 319.53
      float8  : 526.52
      float16 : 116.83

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 82.00
      double2  : 159.28
      double4  : 270.85
      double8  : 56.72
      double16 : 62.36

    Integer compute (GIOPS)
      int   : 202.45
      int2  : 333.43
      int4  : 665.65
      int8  : 1244.93
      int16 : 1437.32

    Integer compute Fast 24bit (GIOPS)
      int   : 121.32
      int2  : 164.70
      int4  : 318.53
      int8  : 583.42
      int16 : 124.19

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer              : 13.87
      enqueueReadBuffer               : 13.96
      enqueueWriteBuffer non-blocking : 13.35
      enqueueReadBuffer non-blocking  : 13.68
      enqueueMapBuffer(for read)      : 4584.72
        memcpy from mapped ptr        : 14.06
      enqueueUnmap(after write)       : 5458.78
        memcpy to mapped ptr          : 14.11

    Kernel launch latency : 185.79 us