1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
|
# ==========================================================================
#
# Copyright NumFOCUS
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0.txt
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# ==========================================================================*/
import itk
width = 256
height = 256
GPUImageType = itk.GPUImage[itk.F, 2]
start = itk.Index[2]()
start[0] = 0
start[1] = 0
size = itk.Size[2]()
size[0] = width
size[1] = height
region = itk.ImageRegion[2](start, size)
srcA = GPUImageType.New()
srcA.SetRegions(region)
srcA.Allocate()
srcA.FillBuffer(1.0)
srcB = GPUImageType.New()
srcB.SetRegions(region)
srcB.Allocate()
srcB.FillBuffer(3.0)
dest = GPUImageType.New()
dest.SetRegions(region)
dest.Allocate()
idx = itk.Index[2]()
idx[0] = 0
idx[1] = 0
number_of_elements = width * height
kernel_manager = itk.GPUKernelManager.New()
gpu_source = itk.GPUImageOps.GetOpenCLSource()
print(gpu_source)
kernel_manager.LoadProgramFromString(gpu_source, "#define PIXELTYPE float\n")
kernel_add = kernel_manager.CreateKernel("ImageAdd")
srcA.SetCurrentCommandQueue(0)
srcB.SetCurrentCommandQueue(0)
dest.SetCurrentCommandQueue(0)
kernel_manager.SetCurrentCommandQueue(0)
print("Current Command Queue ID : 0 ")
print("======================")
print("Kernel : Addition")
print("------------------")
print("Before GPU kernel execution")
print(f"SrcA : {srcA.GetPixel(idx)}")
print(f"SrcB : {srcB.GetPixel(idx)}")
print(f"Dest : {dest.GetPixel(idx)}")
kernel_manager.SetKernelArgWithImage(kernel_add, 0, srcA.GetGPUDataManager())
kernel_manager.SetKernelArgWithImage(kernel_add, 1, srcB.GetGPUDataManager())
kernel_manager.SetKernelArgWithImage(kernel_add, 2, dest.GetGPUDataManager())
kernel_manager.SetKernelArgWithUInt(kernel_add, 3, number_of_elements)
kernel_manager.LaunchKernel2D(kernel_add, width, height, 16, 16)
print("------------------")
print("After GPU kernel execution")
print(f"SrcA : {srcA.GetPixel(idx)}")
print(f"SrcB : {srcB.GetPixel(idx)}")
print(f"Dest : {dest.GetPixel(idx)}")
print("======================")
#
# create multiplication kernel
#
kernel_mult = kernel_manager.CreateKernel("ImageMult")
print("======================")
print("Kernel : Multiplication")
print("------------------")
print("Before GPU kernel execution")
print(f"SrcA : {srcA.GetPixel(idx)}")
print(f"SrcB : {srcB.GetPixel(idx)}")
print(f"Dest : {dest.GetPixel(idx)}")
kernel_manager.SetKernelArgWithImage(kernel_mult, 0, srcA.GetGPUDataManager())
kernel_manager.SetKernelArgWithImage(kernel_mult, 1, srcB.GetGPUDataManager())
kernel_manager.SetKernelArgWithImage(kernel_mult, 2, dest.GetGPUDataManager())
kernel_manager.SetKernelArgWithUInt(kernel_mult, 3, number_of_elements)
kernel_manager.LaunchKernel2D(kernel_mult, width, height, 16, 16)
print("------------------")
print("After GPU kernel execution")
print(f"SrcA : {srcA.GetPixel(idx)}")
print(f"SrcB : {srcB.GetPixel(idx)}")
print(f"Dest : {dest.GetPixel(idx)}")
print("======================")
#
# Change Command Queue if more than one GPU device exists
# otherwise, use same command queue
#
queueID = 0
context_manager = itk.GPUContextManager.GetInstance()
if context_manager.GetNumberOfCommandQueues() >= 2:
queueID = 1
print("More than one GPU device available, switching command queues.")
else:
print("Only one GPU device available, using same command queue.")
print(f"Current Command Queue ID : {queueID}")
#
# create subtraction kernel
#
kernel_sub = kernel_manager.CreateKernel("ImageSub")
srcA.FillBuffer(2.0)
srcB.FillBuffer(4.0)
dest.FillBuffer(1.0)
# default queue id was 0
srcA.SetCurrentCommandQueue(queueID)
srcB.SetCurrentCommandQueue(queueID)
dest.SetCurrentCommandQueue(queueID)
kernel_manager.SetCurrentCommandQueue(queueID)
print("======================")
print("Kernel : Subtraction")
print("------------------")
print("Before GPU kernel execution")
print(f"SrcA : {srcA.GetPixel(idx)}")
print(f"SrcB : {srcB.GetPixel(idx)}")
print(f"Dest : {dest.GetPixel(idx)}")
kernel_manager.SetKernelArgWithImage(kernel_sub, 0, srcA.GetGPUDataManager())
kernel_manager.SetKernelArgWithImage(kernel_sub, 1, srcB.GetGPUDataManager())
kernel_manager.SetKernelArgWithImage(kernel_sub, 2, dest.GetGPUDataManager())
kernel_manager.SetKernelArgWithUInt(kernel_sub, 3, number_of_elements)
kernel_manager.LaunchKernel2D(kernel_sub, width, height, 16, 16)
print("------------------")
print("After GPU kernel execution")
print(f"SrcA : {srcA.GetPixel(idx)}")
print(f"SrcB : {srcB.GetPixel(idx)}")
print(f"Dest : {dest.GetPixel(idx)}")
print("======================")
|
