Geekbench AI Score

17514
Single Precision Score
18969
Half Precision Score
17441
Quantized Score
Geekbench AI 1.1.0 for macOS AArch64

Result Information

Upload Date November 19 2024 10:07 PM
Views 3

System Information

AI Information
Framework Core ML
Backend GPU
Device Apple M3 Max
System Information
Operating System macOS 15.1.1 (Build 24B91)
Model MacBook Pro (14-inch, Nov 2023)
Model IDMac15,8
Motherboard Mac15,8
CPU Information
Name Apple M3 Max
Topology 1 Processor, 16 Cores
Identifier Apple M3 Max
Base Frequency 4.05 GHz
Cluster 112 Cores
Cluster 24 Cores
Memory Information
Size 64.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 7909
1.47 KIPS
 
Image Classification (HP)
100% 8243
1.53 KIPS
 
Image Classification (Q)
97% 7142
1.33 KIPS
 
Image Segmentation (SP)
100% 16500
267.5 IPS
 
Image Segmentation (HP)
100% 17344
281.2 IPS
 
Image Segmentation (Q)
99% 18160
294.4 IPS
 
Pose Estimation (SP)
100% 91896
107.2 IPS
 
Pose Estimation (HP)
100% 95551
111.5 IPS
 
Pose Estimation (Q)
95% 89736
105.2 IPS
 
Object Detection (SP)
100% 7765
616.0 IPS
 
Object Detection (HP)
100% 8308
659.0 IPS
 
Object Detection (Q)
91% 7141
570.7 IPS
 
Face Detection (SP)
100% 31825
378.2 IPS
 
Face Detection (HP)
100% 35981
427.5 IPS
 
Face Detection (Q)
100% 27198
323.2 IPS
 
Depth Estimation (SP)
100% 38436
296.1 IPS
 
Depth Estimation (HP)
99% 40429
311.5 IPS
 
Depth Estimation (Q)
85% 33563
262.3 IPS
 
Style Transfer (SP)
100% 186069
239.2 IPS
 
Style Transfer (HP)
100% 205523
264.2 IPS
 
Style Transfer (Q)
96% 188468
243.2 IPS
 
Image Super-Resolution (SP)
100% 24716
912.6 IPS
 
Image Super-Resolution (HP)
100% 25932
957.5 IPS
 
Image Super-Resolution (Q)
99% 21301
786.5 IPS
 
Text Classification (SP)
100% 2729
3.64 KIPS
 
Text Classification (HP)
100% 4804
6.41 KIPS
 
Text Classification (Q)
93% 4797
6.44 KIPS
 
Machine Translation (SP)
100% 1900
32.7 IPS
 
Machine Translation (HP)
100% 1427
24.6 IPS
 
Machine Translation (Q)
98% 1783
30.8 IPS
 

Compare

Set Baseline Geekbench 6 now available