| Inference Score |
3631 |
1656 |
219.3% |
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Classification (F32)
|
2595
|
935
|
277.5%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Classification (F16)
|
1877
|
1004
|
187.0%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Classification (I8)
|
1853
|
1334
|
138.9%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Segmentation (F32)
|
833
|
439
|
189.7%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Segmentation (F16)
|
878
|
437
|
200.9%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Segmentation (I8)
|
898
|
1282
|
70.0%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Pose Estimation (F32)
|
16604
|
2907
|
571.2%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Pose Estimation (F16)
|
18284
|
3565
|
512.9%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Pose Estimation (I8)
|
17983
|
5343
|
336.6%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Object Detection (F32)
|
1505
|
898
|
167.6%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Object Detection (F16)
|
2014
|
952
|
211.6%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Object Detection (I8)
|
1502
|
848
|
177.1%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Face Detection (F32)
|
5935
|
1161
|
511.2%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Face Detection (F16)
|
7207
|
1150
|
626.7%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Face Detection (I8)
|
7569
|
2083
|
363.4%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Depth Estimation (F32)
|
12704
|
4023
|
315.8%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Depth Estimation (F16)
|
13764
|
3961
|
347.5%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Depth Estimation (I8)
|
12682
|
5282
|
240.1%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Style Transfer (F32)
|
28645
|
8905
|
321.7%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Style Transfer (F16)
|
30540
|
8897
|
343.3%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Style Transfer (I8)
|
29685
|
7872
|
377.1%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Super-Resolution (F32)
|
5677
|
1616
|
351.3%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Super-Resolution (F16)
|
5626
|
1606
|
350.3%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Image Super-Resolution (I8)
|
5874
|
1716
|
342.3%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Text Classification (F32)
|
1088
|
1055
|
103.1%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Text Classification (F16)
|
767
|
1058
|
72.5%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Text Classification (I8)
|
715
|
321
|
222.7%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Machine Translation (F32)
|
608
|
1021
|
59.5%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Machine Translation (F16)
|
631
|
1043
|
60.5%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
|
Machine Translation (I8)
|
536
|
1235
|
43.4%
|
|
iPad Pro (12.9-inch 5th generation)
|
|
|
LENOVO 82R4
|
|
