In the first lines of the script, we import all the requested modules.

In order to initialize the context, we use the cl.create_some_context() method. This asks the user which context must be used to perform the calculation:

Choose platform:
[0] <pyopencl.Platform 'NVIDIA CUDA' at 0x1c0a25aecf0>
[1] <pyopencl.Platform 'Intel(R) OpenCL' at 0x1c0a2608400>

Then, we need to instantiate the queue that will receive ElementwiseKernel:

queue = cl.CommandQueue(context)

Input and output vectors are instantiated. The input vectors, vector_a and vector_b, are integer vectors of random values obtained using the random.randint NumPy function. These vectors are then copied into the device by using the PyOpenCL statement:

cl.array_to_device(queue,array)

In ElementwiseKernel, an object is created:

elementwiseSum = cl.elementwise.ElementwiseKernel(context,
               "int *a, int *b, int *c", "c[i] = a[i] + b[i]", "sum")

Finally, we call the elementwiseSum function with the arguments defined previously:

elementwiseSum(vector_a, vector_b, result_vector)

The example ends by printing the input vectors and the result obtained. The output looks like this:

(base) C:>python elementwisePyopencl.py

Choose platform:
[0] <pyopencl.Platform 'NVIDIA CUDA' at 0x1c0a25aecf0>
[1] <pyopencl.Platform 'Intel(R) OpenCL' at 0x1c0a2608400>
Choice [0]:1

Choose device(s):
[0] <pyopencl.Device 'Intel(R) HD Graphics 5500' on 'Intel(R) OpenCL' at 0x1c0a1640db0>
[1] <pyopencl.Device 'Intel(R) Core(TM) i7-5500U CPU @ 2.40GHz' on 'Intel(R) OpenCL' at 0x1c0a15e53f0>
Choice, comma-separated [0]:0
PyOpenCL ELEMENTWISE SUM OF TWO VECTORS
VECTOR LENGTH = 100
INPUT VECTOR A
[24 64 73 37 40 4 41 85 19 90 32 51 6 89 98 56 97 53 34 91 82 89 97 2
 54 65 90 90 91 75 30 8 62 94 63 69 31 99 8 18 28 7 81 72 14 53 91 80
 76 39 8 47 25 45 26 56 23 47 41 18 89 17 82 84 10 75 56 89 71 56 66 61
 58 54 27 88 16 20 9 61 68 63 74 84 18 82 67 30 15 25 25 3 93 36 24 27
 70 5 78 15]

INPUT VECTOR B
[49 18 69 43 51 72 37 50 79 34 97 49 51 29 89 81 33 7 47 93 70 52 63 90
 99 95 58 33 41 70 84 87 20 83 74 43 78 34 94 47 89 4 30 36 34 56 32 31
 56 22 50 52 68 98 52 80 14 98 43 60 20 49 15 38 74 89 99 29 96 65 89 41
 72 53 89 31 34 64 0 47 87 70 98 86 41 25 34 10 44 36 54 52 54 86 33 38
 25 49 75 53]

OUTPUT VECTOR RESULT A + B
[73 82 142 80 91 76 78 135 98 124 129 100 57 118 187 137 130 60 81 184 
 152 141 160 92 153 160 148 123 132 145 114 95 82 177 137 112 109 133 
 102 65 117 11 111 108 48 109 123 111 132 61 58 99 93 143 78 136 37 145 
 84 78 109 66 97 122 84 164 155 118 167 121 155 102 130 107 116 119 50 
 84 9 108 155 133 172 170 59 107 101 40 59 61 79 55 147 122 57 65 
 95 54 153 68]