Platform: Intel(R) OpenCL Graphics
  Device: Intel(R) Iris(R) Xe Graphics
    Driver version  : 23.43.027642 (Linux x64)
    Compute units   : 96
    Clock frequency : 1500 MHz

    Global memory bandwidth (GBPS)
      float   : 72.06
      float2  : 73.81
      float4  : 75.24
      float8  : 74.04
      float16 : 71.80

    Single-precision compute (GFLOPS)
      float   : 2246.59
      float2  : 2219.88
      float4  : 2236.12
      float8  : 1420.12
      float16 : 1409.69

    Half-precision compute (GFLOPS)
      half   : 4356.72
      half2  : 4315.23
      half4  : 4355.17
      half8  : 4326.94
      half16 : 3939.02

    No double precision support! Skipped

    Integer compute (GIOPS)
      int   : 757.15
      int2  : 511.17
      int4  : 488.24
      int8  : 464.11
      int16 : 563.47

    Integer compute Fast 24bit (GIOPS)
      int   : 753.83
      int2  : 511.19
      int4  : 488.22
      int8  : 464.11
      int16 : 563.45

    Integer char (8bit) compute (GIOPS)
      char   : 1494.68
      char2  : 1483.94
      char4  : 1470.63
      char8  : 1393.09
      char16 : 1368.20

    Integer short (16bit) compute (GIOPS)
      short   : 4239.22
      short2  : 4164.17
      short4  : 4207.74
      short8  : 4154.31
      short16 : 4073.11

    Transfer bandwidth (GBPS)
      enqueueWriteBuffer              : 20.67
      enqueueReadBuffer               : 20.93
      enqueueWriteBuffer non-blocking : 31.51
      enqueueReadBuffer non-blocking  : 32.88
      enqueueMapBuffer(for read)      : 7158265.00
        memcpy from mapped ptr        : 20.92
      enqueueUnmap(after write)       : inf
        memcpy to mapped ptr          : 20.64

    Kernel launch latency : 34.44 us