Upload Date | September 18 2024 10:12 AM |
Views | 1 |
AI Information | |
---|---|
Framework | Core ML |
Backend | Neural Engine |
Device | Apple M1 |
System Information | |
---|---|
Operating System | iOS 18.1 |
Model | iPad Pro (11-inch 3rd generation) |
Model ID | iPad13,5 |
Motherboard | J517xAP |
CPU Information | |
---|---|
Name | Apple M1 |
Topology | 1 Processor, 8 Cores |
Identifier | ARM |
Base Frequency | 3.19 GHz |
Cluster 1 | 4 Cores |
Cluster 2 | 4 Cores |
Memory Information | |
---|---|
Size | 15.37 GB |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (SP)
|
100% |
1400
260.4 IPS |
|
Image Classification (HP)
|
100% |
6075
1.13 KIPS |
|
Image Classification (Q)
|
96% |
5865
1.10 KIPS |
|
Image Segmentation (SP)
|
100% |
1429
23.2 IPS |
|
Image Segmentation (HP)
|
100% |
6626
107.4 IPS |
|
Image Segmentation (Q)
|
99% |
6753
109.5 IPS |
|
Pose Estimation (SP)
|
100% |
6523
7.61 IPS |
|
Pose Estimation (HP)
|
100% |
68996
80.5 IPS |
|
Pose Estimation (Q)
|
95% |
66545
78.0 IPS |
|
Object Detection (SP)
|
100% |
1515
120.1 IPS |
|
Object Detection (HP)
|
99% |
7329
583.0 IPS |
|
Object Detection (Q)
|
91% |
6621
529.0 IPS |
|
Face Detection (SP)
|
100% |
3037
36.1 IPS |
|
Face Detection (HP)
|
100% |
16145
191.8 IPS |
|
Face Detection (Q)
|
100% |
14883
176.8 IPS |
|
Depth Estimation (SP)
|
100% |
5808
44.7 IPS |
|
Depth Estimation (HP)
|
99% |
48149
371.0 IPS |
|
Depth Estimation (Q)
|
85% |
46881
366.4 IPS |
|
Style Transfer (SP)
|
100% |
17267
22.2 IPS |
|
Style Transfer (HP)
|
100% |
118959
152.9 IPS |
|
Style Transfer (Q)
|
96% |
111130
143.4 IPS |
|
Image Super-Resolution (SP)
|
100% |
3806
140.5 IPS |
|
Image Super-Resolution (HP)
|
100% |
22271
822.4 IPS |
|
Image Super-Resolution (Q)
|
99% |
23001
849.3 IPS |
|
Text Classification (SP)
|
100% |
2307
3.08 KIPS |
|
Text Classification (HP)
|
100% |
4017
5.36 KIPS |
|
Text Classification (Q)
|
93% |
4001
5.37 KIPS |
|
Machine Translation (SP)
|
100% |
2589
44.6 IPS |
|
Machine Translation (HP)
|
100% |
2328
40.1 IPS |
|
Machine Translation (Q)
|
98% |
2325
40.2 IPS |