Platform: Portable Computing Language
  Device: Xavier
    Driver version  : 3.0-rc2 (Linux ARM64)
    Compute units   : 8
    Clock frequency : 1377 MHz

    Global memory bandwidth (GBPS)
      float   : 109.54
      float2  : 108.59
      float4  : 108.73
      float8  : 91.07
      float16 : 105.74

    Single-precision compute (GFLOPS)
      float   : 1400.08
      float2  : 1403.62
      float4  : 1400.64
      float8  : 1395.61
      float16 : 1385.01

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 44.01
      double2  : 43.95
      double4  : 43.84
      double8  : 43.53
      double16 : 43.06

    Integer compute (GIOPS)
      int   : 1397.38
      int2  : 1401.00
      int4  : 1394.18
      int8  : 1400.12
      int16 : 1400.80

    Integer compute Fast 24bit (GIOPS)
      int   : 1397.19
      int2  : 1400.98
      int4  : 1393.72
      int8  : 1400.03
      int16 : 1398.92

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer              : 10.50
      enqueueReadBuffer               : 10.77
      enqueueWriteBuffer non-blocking : 10.53
      enqueueReadBuffer non-blocking  : 10.75
      enqueueMapBuffer(for read)      : 72523.52
        memcpy from mapped ptr        : 10.72
      enqueueUnmap(after write)       : 14.82
        memcpy to mapped ptr          : 10.55

    Kernel launch latency : -44.15 us