Geekbench AI Score

498
Single Precision Score
309
Half Precision Score
421
Quantized Score
Geekbench AI 1.2.0 for Windows AVX2

Result Information

Upload Date December 27 2024 11:47 PM
Views 1

System Information

AI Information
Framework ONNX
Backend DirectML
Device Intel(R) HD Graphics 630
System Information
Operating System Microsoft Windows 10 Enterprise 2016 LTSB (64-bit)
Model Dell Inc. Precision 7520
Motherboard Dell Inc. 0P1FNM
Power PlanHigh performance
CPU Information
Name Intel Core i7-7820HQ
Topology 1 Processor, 4 Cores, 8 Threads
Identifier GenuineIntel Family 6 Model 158 Stepping 9
Base Frequency 2.90 GHz
Cluster 14 Cores
Memory Information
Size 16.00 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 272
50.7 IPS
 
Image Classification (HP)
100% 159
29.5 IPS
 
Image Classification (Q)
100% 257
47.7 IPS
 
Image Segmentation (SP)
100% 392
6.36 IPS
 
Image Segmentation (HP)
100% 219
3.56 IPS
 
Image Segmentation (Q)
98% 340
5.53 IPS
 
Pose Estimation (SP)
100% 883
1.03 IPS
 
Pose Estimation (HP)
100% 592
0.69 IPS
 
Pose Estimation (Q)
97% 846
0.99 IPS
 
Object Detection (SP)
100% 294
23.3 IPS
 
Object Detection (HP)
100% 172
13.6 IPS
 
Object Detection (Q)
86% 263
21.2 IPS
 
Face Detection (SP)
100% 611
7.26 IPS
 
Face Detection (HP)
100% 352
4.19 IPS
 
Face Detection (Q)
97% 511
6.09 IPS
 
Depth Estimation (SP)
100% 844
6.51 IPS
 
Depth Estimation (HP)
99% 548
4.23 IPS
 
Depth Estimation (Q)
77% 766
6.14 IPS
 
Style Transfer (SP)
100% 1756
2.26 IPS
 
Style Transfer (HP)
100% 1298
1.67 IPS
 
Style Transfer (Q)
98% 1555
2.00 IPS
 
Image Super-Resolution (SP)
100% 466
17.2 IPS
 
Image Super-Resolution (HP)
100% 301
11.1 IPS
 
Image Super-Resolution (Q)
99% 419
15.5 IPS
 
Text Classification (SP)
100% 248
331.1 IPS
 
Text Classification (HP)
100% 136
181.4 IPS
 
Text Classification (Q)
97% 176
235.9 IPS
 
Machine Translation (SP)
100% 329
5.66 IPS
 
Machine Translation (HP)
96% 224
3.87 IPS
 
Machine Translation (Q)
70% 205
3.84 IPS
 

Compare

Set Baseline Geekbench 6 now available