Platform: NVIDIA CUDA
  Device: GeForce GTX 750
    Driver version  : 343.22 (Linux x64)
    Compute units   : 4
    Clock frequency : 1137 MHz

    Global memory bandwidth (GBPS)
      float   : 68.31
      float2  : 71.03
      float4  : 72.28
      float8  : 68.97
      float16 : 65.70

    Single-precision compute (GFLOPS)
      float   : 750.57
      float2  : 962.20
      float4  : 1172.67
      float8  : 1177.41
      float16 : 1168.44

    Double-precision compute (GFLOPS)
      double   : 38.06
      double2  : 38.03
      double4  : 37.97
      double8  : 37.82
      double16 : 37.55

    Integer compute (GIOPS)
      int   : 273.43
      int2  : 305.58
      int4  : 333.16
      int8  : 332.69
      int16 : 340.65

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 7.97
      enqueueReadBuffer          : 4.65
      enqueueMapBuffer(for read) : 8.75
        memcpy from mapped ptr   : 3.58
      enqueueUnmap(after write)  : 12.40
        memcpy to mapped ptr     : 3.50

    Kernel launch latency : 4.60 us