User | spotlite |
Upload Date | December 05 2024 10:01 PM |
Views | 6 |
AI Information | |
---|---|
Framework | Core ML |
Backend | CPU |
Device | Apple A15 Bionic |
System Information | |
---|---|
Operating System | iOS 18.1.1 |
Model | iPad mini (6th generation) |
Model ID | iPad14,1 |
Motherboard | J310AP |
CPU Information | |
---|---|
Name | Apple A15 Bionic |
Topology | 1 Processor, 6 Cores |
Identifier | ARM |
Base Frequency | 2.92 GHz |
Cluster 1 | 2 Cores |
Cluster 2 | 4 Cores |
Memory Information | |
---|---|
Size | 3.77 GB |
Workload | Accuracy | Score | |
---|---|---|---|
Image Classification (SP)
|
100% |
1304
242.6 IPS |
|
Image Classification (HP)
|
100% |
2246
417.7 IPS |
|
Image Classification (Q)
|
96% |
1552
289.8 IPS |
|
Image Segmentation (SP)
|
100% |
1278
20.7 IPS |
|
Image Segmentation (HP)
|
100% |
1906
30.9 IPS |
|
Image Segmentation (Q)
|
99% |
2352
38.1 IPS |
|
Pose Estimation (SP)
|
100% |
6107
7.13 IPS |
|
Pose Estimation (HP)
|
100% |
13237
15.4 IPS |
|
Pose Estimation (Q)
|
95% |
11526
13.5 IPS |
|
Object Detection (SP)
|
100% |
1391
110.3 IPS |
|
Object Detection (HP)
|
97% |
2621
208.7 IPS |
|
Object Detection (Q)
|
94% |
1853
147.7 IPS |
|
Face Detection (SP)
|
100% |
2682
31.9 IPS |
|
Face Detection (HP)
|
100% |
5500
65.4 IPS |
|
Face Detection (Q)
|
100% |
3038
36.1 IPS |
|
Depth Estimation (SP)
|
100% |
5248
40.4 IPS |
|
Depth Estimation (HP)
|
98% |
10731
82.9 IPS |
|
Depth Estimation (Q)
|
85% |
7252
56.7 IPS |
|
Style Transfer (SP)
|
100% |
16880
21.7 IPS |
|
Style Transfer (HP)
|
100% |
25713
33.1 IPS |
|
Style Transfer (Q)
|
96% |
21273
27.5 IPS |
|
Image Super-Resolution (SP)
|
100% |
3816
140.9 IPS |
|
Image Super-Resolution (HP)
|
100% |
4525
167.1 IPS |
|
Image Super-Resolution (Q)
|
99% |
3383
124.9 IPS |
|
Text Classification (SP)
|
100% |
2419
3.23 KIPS |
|
Text Classification (HP)
|
100% |
4201
5.61 KIPS |
|
Text Classification (Q)
|
93% |
4148
5.57 KIPS |
|
Machine Translation (SP)
|
100% |
3357
57.8 IPS |
|
Machine Translation (HP)
|
98% |
5507
95.2 IPS |
|
Machine Translation (Q)
|
98% |
5500
95.0 IPS |