Geekbench AI Score

2896
Single Precision Score
882
Half Precision Score
7053
Quantized Score
Geekbench AI 1.7.0 for Linux AVX2

Result Information

User diegotomc
Upload Date March 27 2026 04:03 AM
Views 7

System Information

AI Information
Framework ONNX
Backend CPU
Device Intel(R) Core(TM) Ultra 9 288V
System Information
Operating System Fedora Linux 43 (Workstation Edition)
Model Acer Swift SFE14-51T
Motherboard LNL Sunfish_LNV
CPU Information
Name Intel(R) Core(TM) Ultra 9 288V
Topology 1 Processor, 8 Cores
Identifier GenuineIntel Family 6 Model 189 Stepping 1
Base Frequency 5.00 GHz
Cluster 10 Cores
Memory Information
Size 30.91 GB

AI Performance

Workload Accuracy Score
Image Classification (SP)
100% 1715
319.0 IPS
 
Image Classification (HP)
100% 221
41.1 IPS
 
Image Classification (Q)
97% 6168
1.15 KIPS
 
Image Segmentation (SP)
100% 1480
24.0 IPS
 
Image Segmentation (HP)
100% 372
6.03 IPS
 
Image Segmentation (Q)
99% 3333
54.2 IPS
 
Pose Estimation (SP)
100% 4522
5.28 IPS
 
Pose Estimation (HP)
100% 3348
3.91 IPS
 
Pose Estimation (Q)
94% 18234
21.4 IPS
 
Object Detection (SP)
100% 1831
145.2 IPS
 
Object Detection (HP)
100% 291
23.1 IPS
 
Object Detection (Q)
86% 5813
467.2 IPS
 
Face Detection (SP)
100% 3732
44.3 IPS
 
Face Detection (HP)
100% 445
5.29 IPS
 
Face Detection (Q)
97% 14813
176.6 IPS
 
Depth Estimation (SP)
100% 5772
44.5 IPS
 
Depth Estimation (HP)
99% 1235
9.51 IPS
 
Depth Estimation (Q)
78% 13905
111.0 IPS
 
Style Transfer (SP)
100% 11505
14.8 IPS
 
Style Transfer (HP)
100% 8730
11.2 IPS
 
Style Transfer (Q)
98% 27213
35.1 IPS
 
Image Super-Resolution (SP)
100% 2407
88.9 IPS
 
Image Super-Resolution (HP)
100% 1738
64.2 IPS
 
Image Super-Resolution (Q)
99% 4596
170.2 IPS
 
Text Classification (SP)
100% 1459
1.95 KIPS
 
Text Classification (HP)
100% 624
833.3 IPS
 
Text Classification (Q)
97% 1712
2.29 KIPS
 
Machine Translation (SP)
100% 2275
39.2 IPS
 
Machine Translation (HP)
100% 690
11.9 IPS
 
Machine Translation (Q)
65% 3174
63.8 IPS
 

Compare

Set Baseline Geekbench 6 now available