OpenCL performance on intel Core i3-8100 by Clpeak

Platform: Intel(R) CPU Runtime for OpenCL(TM) Applications
  Device: Intel(R) Core(TM) i3-8100 CPU @ 3.60GHz
    Driver version  : 18.1.0.0920 (Linux x64)
    Compute units   : 4
    Clock frequency : 3600 MHz

    Global memory bandwidth (GBPS)
      float   : 14.32
      float2  : 15.14
      float4  : 15.36
      float8  : 14.86
      float16 : 12.22

    Single-precision compute (GFLOPS)
      float   : 58.82
      float2  : 117.37
      float4  : 232.28
      float8  : 452.27
      float16 : 130.48

    No half precision support! Skipped

    Double-precision compute (GFLOPS)
      double   : 57.94
      double2  : 115.19
      double4  : 226.52
      double8  : 131.14
      double16 : 117.21

    Integer compute (GIOPS)
      int   : 20.95
      int2  : 41.47
      int4  : 81.03
      int8  : 57.61
      int16 : 143.54

    Integer compute Fast 24bit (GIOPS)
      int   : 17.82
      int2  : 35.46
      int4  : 65.71
      int8  : 97.62
      int16 : 36.05

    Integer char (8bit) compute (GIOPS)
      char   : 36.80
      char2  : 71.86
      char4  : 139.64
      char8  : 46.01
      char16 : 63.98

    Integer short (16bit) compute (GIOPS)
      short   : 36.85
      short2  : 71.95
      short4  : 139.49
      short8  : 46.57
      short16 : 65.29

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer              : 6.92
      enqueueReadBuffer               : 7.11
      enqueueWriteBuffer non-blocking : 6.92
      enqueueReadBuffer non-blocking  : 7.12
      enqueueMapBuffer(for read)      : 78375.32
        memcpy from mapped ptr        : 7.10
      enqueueUnmap(after write)       : 77247.62
        memcpy to mapped ptr          : 6.92

    Kernel launch latency : 1.59 us