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
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
|
Identifier: "Occupancy"
DisplayName: "Occupancy"
Description: "Occupancy is the ratio of the number of active warps per multiprocessor to the maximum number of possible active warps. Another way to view occupancy is the percentage of the hardware's ability to process warps that is actively in use. Higher occupancy does not always result in higher performance, however, low occupancy always reduces the ability to hide latencies, resulting in overall performance degradation. Large discrepancies between the theoretical and the achieved occupancy during execution typically indicates highly imbalanced workloads."
Order: 80
Sets {
Identifier: "basic"
}
Sets {
Identifier: "detailed"
}
Sets {
Identifier: "full"
}
Filter {
CollectionFilter {
CollectionScopes: CollectionScope_Launch
}
}
Metrics {
Metrics {
Label: "GPU Maximum Warps Per Scheduler"
Name: "smsp__warps_active.avg.peak_sustained"
}
Metrics {
Label: "Theoretical Warps Per Scheduler"
Name: "smsp__maximum_warps_avg_per_active_cycle"
}
}
Header {
Metrics {
Label: "Theoretical Occupancy"
Name: "sm__maximum_warps_per_active_cycle_pct"
}
Metrics {
Label: "Block Limit Registers"
Name: "launch__occupancy_limit_registers"
}
Metrics {
Label: "Theoretical Active Warps per SM"
Name: "sm__maximum_warps_avg_per_active_cycle"
}
Metrics {
Label: "Block Limit Shared Mem"
Name: "launch__occupancy_limit_shared_mem"
}
Metrics {
Label: "Achieved Occupancy"
Name: "sm__warps_active.avg.pct_of_peak_sustained_active"
Filter {
CollectionFilter {
CollectionScopes: CollectionScope_Launch
CollectionScopes: CollectionScope_Range
}
}
}
Metrics {
Label: "Block Limit Warps"
Name: "launch__occupancy_limit_warps"
}
Metrics {
Label: "Achieved Active Warps Per SM"
Name: "sm__warps_active.avg.per_cycle_active"
Filter {
CollectionFilter {
CollectionScopes: CollectionScope_Launch
CollectionScopes: CollectionScope_Range
}
}
}
Metrics {
Label: "Block Limit SM"
Name: "launch__occupancy_limit_blocks"
}
Metrics {
Label: "Max Cluster Size"
Name: "launch__cluster_max_potential_size"
Filter {
MinArch: CC_90
}
}
Metrics {
Label: "Max Active Clusters"
Name: "launch__cluster_max_active"
Filter {
MinArch: CC_90
}
}
Metrics {
Label: "Cluster Occupancy"
Name: "launch__occupancy_cluster_pct"
Filter {
MinArch: CC_90
}
}
Metrics {
Label: "Overall GPU Occupancy"
Name: "launch__occupancy_cluster_gpu_pct"
Filter {
MinArch: CC_90
}
}
}
Body {
Items {
LineChart {
Label: "Impact of Varying Register Count Per Thread"
AxisX {
Label: "Registers Per Thread"
Range {
Min: 0
}
TickCount: 33
Precision: 0
}
AxisY {
Label: "Warp Occupancy"
Range {
Min: 0
}
Precision: 0
}
Metrics {
Label: "Warp Occupancy Per Register Count"
Name: "launch__occupancy_per_register_count"
ShowInstances: 1
}
HighlightX {
Metrics {
Label: "Registers Per Thread"
Name: "launch__registers_per_thread"
}
}
}
}
Items {
LineChart {
Label: "Impact of Varying Block Size"
AxisX {
Label: "Block Size"
Range {
Min: 0
}
TickCount: 33
Precision: 0
}
AxisY {
Label: "Warp Occupancy"
Range {
Min: 0
}
Precision: 0
}
Metrics {
Label: "Warp Occupancy Per Block Size"
Name: "launch__occupancy_per_block_size"
ShowInstances: 1
}
HighlightX {
Metrics {
Label: "Block Size"
Name: "launch__block_size"
}
}
}
}
Items {
LineChart {
Label: "Impact of Varying Shared Memory Usage Per Block"
AxisX {
Label: "Shared Memory Per Block"
Range {
Min: 0
}
TickCount: 33
Precision: 0
}
AxisY {
Label: "Warp Occupancy"
Range {
Min: 0
}
Precision: 0
}
Metrics {
Label: "Warp Occupancy Per Shared Mem Size"
Name: "launch__occupancy_per_shared_mem_size"
ShowInstances: 1
}
HighlightX {
Metrics {
Label: "Shared Memory Per Block"
Name: "launch__shared_mem_per_block"
}
}
}
}
Items {
LineChart {
Label: "Impact of Varying Cluster Size"
AxisX {
Label: "Potential Cluster Size"
Range {
Min: 1
Max: 16
}
TickCount: 16
Precision: 0
}
AxisY {
Label: "Active Clusters"
Range {
Min: 0
}
Precision: 0
}
Metrics {
Label: "Active Clusters"
Name: "launch__occupancy_per_cluster_size"
ShowInstances: 1
}
HighlightX {
Metrics {
Label: "Cluster Size"
Name: "launch__cluster_size"
}
}
}
Filter {
MinArch: CC_90
}
}
}
|
