User | ramphex |
Upload Date | November 12 2024 07:28 AM |
Views | 6 |
AI Information | |
---|---|
Framework | Core ML |
Backend | GPU |
Device | Apple M2 |
System Information | |
---|---|
Operating System | iOS 18.2 |
Model | iPad Pro (11-inch, 4th generation) |
Model ID | iPad14,3 |
Motherboard | J617AP |
CPU Information | |
---|---|
Name | Apple M2 |
Topology | 1 Processor, 8 Cores |
Identifier | ARM |
Base Frequency | 3.49 GHz |
Cluster 1 | 4 Cores |
Cluster 2 | 4 Cores |
Memory Information | |
---|---|
Size | 7.37 GB |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (SP)
|
100% |
3581
665.9 IPS |
|
Image Classification (HP)
|
100% |
3569
663.7 IPS |
|
Image Classification (Q)
|
97% |
2983
556.8 IPS |
|
Image Segmentation (SP)
|
100% |
4058
65.8 IPS |
|
Image Segmentation (HP)
|
100% |
5017
81.3 IPS |
|
Image Segmentation (Q)
|
99% |
5115
82.9 IPS |
|
Pose Estimation (SP)
|
100% |
24734
28.9 IPS |
|
Pose Estimation (HP)
|
100% |
26258
30.6 IPS |
|
Pose Estimation (Q)
|
95% |
24537
28.8 IPS |
|
Object Detection (SP)
|
100% |
3647
289.3 IPS |
|
Object Detection (HP)
|
100% |
3716
294.8 IPS |
|
Object Detection (Q)
|
91% |
3065
244.9 IPS |
|
Face Detection (SP)
|
100% |
10405
123.6 IPS |
|
Face Detection (HP)
|
100% |
11996
142.5 IPS |
|
Face Detection (Q)
|
100% |
8357
99.3 IPS |
|
Depth Estimation (SP)
|
100% |
16990
130.9 IPS |
|
Depth Estimation (HP)
|
99% |
17979
138.5 IPS |
|
Depth Estimation (Q)
|
85% |
14944
116.8 IPS |
|
Style Transfer (SP)
|
100% |
43106
55.4 IPS |
|
Style Transfer (HP)
|
100% |
48532
62.4 IPS |
|
Style Transfer (Q)
|
96% |
45482
58.7 IPS |
|
Image Super-Resolution (SP)
|
100% |
9813
362.4 IPS |
|
Image Super-Resolution (HP)
|
100% |
10053
371.2 IPS |
|
Image Super-Resolution (Q)
|
99% |
7977
294.6 IPS |
|
Text Classification (SP)
|
100% |
2199
2.93 KIPS |
|
Text Classification (HP)
|
100% |
3756
5.01 KIPS |
|
Text Classification (Q)
|
93% |
3747
5.03 KIPS |
|
Machine Translation (SP)
|
100% |
1253
21.6 IPS |
|
Machine Translation (HP)
|
100% |
1267
21.8 IPS |
|
Machine Translation (Q)
|
98% |
1246
21.5 IPS |