GeForce_RTX_2080_Super.log

Platform: NVIDIA CUDA
  Device: GeForce RTX 2080 SUPER
    Driver version  : 440.64 (Linux x64)
    Compute units   : 48
    Clock frequency : 1830 MHz

    Global memory bandwidth (GBPS)
      float   : 379.44
      float2  : 400.04
      float4  : 406.76
      float8  : 421.57
      float16 : 387.33

    Single-precision compute (GFLOPS)
      float   : 11928.26
      float2  : 11929.92
      float4  : 11826.57
      float8  : 11323.94
      float16 : 11032.11

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 368.15
      double2  : 362.51
      double4  : 361.13
      double8  : 350.26
      double16 : 356.73

    Integer compute (GIOPS)
      int   : 10998.23
      int2  : 11085.42
      int4  : 11515.90
      int8  : 11564.03
      int16 : 11580.59

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer         : 9.02
      enqueueReadBuffer          : 8.65
      enqueueMapBuffer(for read) : 11.40
        memcpy from mapped ptr   : 11.83
      enqueueUnmap(after write)  : 12.93
        memcpy to mapped ptr     : 11.99

    Kernel launch latency : 3.46 us