Geekbench AI Score

3374
Single Precision Score
833
Half Precision Score
7219
Quantized Score
Geekbench AI 1.5.0 for Linux AVX2

Result Information

User ce107
Upload Date November 28 2025 06:16 PM
Views 4

System Information

AI Information
Framework ONNX
Backend CPU
Device Intel Core i7-13850HX
System Information
Operating System Fedora Linux 42 (Workstation Edition)
Model Dell Inc. Precision 7680
Motherboard Dell Inc. 0XRC2C
CPU Information
Name Intel Core i7-13850HX
Topology 1 Processor, 20 Cores, 28 Threads
Identifier GenuineIntel Family 6 Model 183 Stepping 1
Base Frequency 5.10 GHz
Cluster 10 Cores
Memory Information
Size 62.47 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 1992
370.4 IPS
 
Image Classification (HP)
100% 207
38.5 IPS
 
Image Classification (Q)
97% 6505
1.21 KIPS
 
Image Segmentation (SP)
100% 1540
25.0 IPS
 
Image Segmentation (HP)
100% 352
5.71 IPS
 
Image Segmentation (Q)
99% 3414
55.5 IPS
 
Pose Estimation (SP)
100% 6263
7.31 IPS
 
Pose Estimation (HP)
100% 3576
4.17 IPS
 
Pose Estimation (Q)
94% 25647
30.1 IPS
 
Object Detection (SP)
100% 2152
170.7 IPS
 
Object Detection (HP)
100% 276
21.9 IPS
 
Object Detection (Q)
86% 6213
499.3 IPS
 
Face Detection (SP)
100% 3822
45.4 IPS
 
Face Detection (HP)
100% 434
5.16 IPS
 
Face Detection (Q)
97% 14492
172.8 IPS
 
Depth Estimation (SP)
100% 6910
53.2 IPS
 
Depth Estimation (HP)
99% 1035
7.97 IPS
 
Depth Estimation (Q)
78% 13536
108.0 IPS
 
Style Transfer (SP)
100% 15360
19.7 IPS
 
Style Transfer (HP)
100% 9050
11.6 IPS
 
Style Transfer (Q)
98% 26371
34.0 IPS
 
Image Super-Resolution (SP)
100% 2738
101.1 IPS
 
Image Super-Resolution (HP)
100% 1776
65.6 IPS
 
Image Super-Resolution (Q)
99% 4609
170.7 IPS
 
Text Classification (SP)
100% 1643
2.19 KIPS
 
Text Classification (HP)
100% 573
764.6 IPS
 
Text Classification (Q)
97% 1524
2.04 KIPS
 
Machine Translation (SP)
100% 2538
43.7 IPS
 
Machine Translation (HP)
100% 544
9.37 IPS
 
Machine Translation (Q)
65% 2991
60.1 IPS
 

Compare

Set Baseline Geekbench 6 now available