Platform: Portable Computing Language
  Device: Quadro K620
    Driver version  : 3.0-rc2 (Linux x64)
    Compute units   : 3
    Clock frequency : 1124 MHz

    Global memory bandwidth (GBPS)
      float   : 25.40
      float2  : 26.21
      float4  : 26.68
      float8  : 25.46
      float16 : 25.53

    Single-precision compute (GFLOPS)
      float   : 572.16
      float2  : 849.36
      float4  : 862.07
      float8  : 807.75
      float16 : 840.97

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 27.51
      double2  : 27.48
      double4  : 27.44
      double8  : 27.33
      double16 : 27.13

    Integer compute (GIOPS)
      int   : 247.08
      int2  : 282.05
      int4  : 289.38
      int8  : 274.71
      int16 : 263.76

    Integer compute Fast 24bit (GIOPS)
      int   : 246.85
      int2  : 282.00
      int4  : 289.37
      int8  : 275.08
      int16 : 264.50

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer              : 5.52
      enqueueReadBuffer               : 5.38
      enqueueWriteBuffer non-blocking : 5.52
      enqueueReadBuffer non-blocking  : 5.38
      enqueueMapBuffer(for read)      : 13252.01
        memcpy from mapped ptr        : 4.43
      enqueueUnmap(after write)       : 4.71
        memcpy to mapped ptr          : 4.52

    Kernel launch latency : -3968.82 us