Platform: NVIDIA CUDA
  Device: GeForce GT 630
    Driver version  : 337.88 (Win32)
    Compute units   : 2
    Clock frequency : 1620 MHz

    Global memory bandwidth (GBPS)
      float   : 17.03
      float2  : 17.43
      float4  : 17.66
      float8  : 8.83
      float16 : 4.55

    Single-precision compute (GFLOPS)
      float   : 206.15
      float2  : 308.23
      float4  : 307.50
      float8  : 294.83
      float16 : 279.48

    Double-precision compute (GFLOPS)
      double   : 25.89
      double2  : 25.87
      double4  : 25.81
      double8  : 25.72
      double16 : 25.52

    Integer compute (GIOPS)
      int   : 103.40
      int2  : 103.49
      int4  : 103.40
      int8  : 103.40
      int16 : 103.40

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 3.11
      enqueueReadBuffer          : 2.99
      enqueueMapBuffer(for read) : 2.91
        memcpy from mapped ptr   : 3.75
      enqueueUnmap(after write)  : 4.17
        memcpy to mapped ptr     : 3.78

    Kernel launch latency : 23.75 us