Geekbench AI Score

812
Single Precision Score
811
Half Precision Score
1118
Quantized Score
Geekbench AI 1.3.0 for Linux x86 (64-bit)

Result Information

User SiggiTron
Upload Date April 06 2025 01:13 PM
Views 2

System Information

AI Information
Framework OpenVINO
Backend CPU
Device Intel(R) Xeon(R) CPU E5-2643 0 @ 3.30GHz
System Information
Operating System Ubuntu 24.04.2 LTS
Model Hewlett-Packard HP Z820 Workstation
Motherboard Hewlett-Packard 158B
CPU Information
Name Intel Xeon E5-2643
Topology 2 Processors, 8 Cores, 16 Threads
Identifier GenuineIntel Family 6 Model 45 Stepping 7
Base Frequency 3.50 GHz
Cluster 10 Cores
Memory Information
Size 251.82 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 405
75.3 IPS
 
Image Classification (HP)
100% 405
75.3 IPS
 
Image Classification (Q)
100% 717
133.4 IPS
 
Image Segmentation (SP)
100% 662
10.7 IPS
 
Image Segmentation (HP)
100% 659
10.7 IPS
 
Image Segmentation (Q)
99% 860
14.0 IPS
 
Pose Estimation (SP)
100% 697
0.81 IPS
 
Pose Estimation (HP)
100% 697
0.81 IPS
 
Pose Estimation (Q)
94% 1503
1.76 IPS
 
Object Detection (SP)
100% 411
32.6 IPS
 
Object Detection (HP)
100% 410
32.5 IPS
 
Object Detection (Q)
87% 703
56.4 IPS
 
Face Detection (SP)
100% 1454
17.3 IPS
 
Face Detection (HP)
100% 1450
17.2 IPS
 
Face Detection (Q)
97% 2004
23.9 IPS
 
Depth Estimation (SP)
100% 1369
10.5 IPS
 
Depth Estimation (HP)
99% 1371
10.6 IPS
 
Depth Estimation (Q)
81% 1852
14.6 IPS
 
Style Transfer (SP)
100% 2160
2.78 IPS
 
Style Transfer (HP)
100% 2157
2.77 IPS
 
Style Transfer (Q)
98% 4533
5.84 IPS
 
Image Super-Resolution (SP)
100% 628
23.2 IPS
 
Image Super-Resolution (HP)
100% 628
23.2 IPS
 
Image Super-Resolution (Q)
98% 911
33.8 IPS
 
Text Classification (SP)
100% 692
924.3 IPS
 
Text Classification (HP)
100% 696
928.4 IPS
 
Text Classification (Q)
91% 354
475.2 IPS
 
Machine Translation (SP)
100% 875
15.1 IPS
 
Machine Translation (HP)
100% 863
14.9 IPS
 
Machine Translation (Q)
100% 864
14.9 IPS
 

Compare

Set Baseline Geekbench 6 now available