Geekbench ML Score

2484
ONNX DirectML Inference Score
Geekbench ML 0.6.0 for Windows AVX2

Result Information

Upload Date August 19 2024 11:43 PM
Views 2

System Information

System Information
Operating System Microsoft Windows 11 Pro (64-bit)
Model LENOVO 21KWZC45US
Motherboard LENOVO 21KWZC45US
Power PlanBalanced
CPU Information
Name Intel Core Ultra 5 135H
Topology 1 Processor, 14 Cores, 18 Threads
Identifier GenuineIntel Family 6 Model 170 Stepping 4
Base Frequency 1.70 GHz
Cluster 14 Cores
Cluster 210 Cores
L1 Instruction Cache64.0 KB x 18
L1 Data Cache32.0 KB x 18
L2 Cache2.00 MB x 2
L3 Cache18.0 MB x 1
Memory Information
Size 16.00 GB
Inference Information
Framework ONNX
Backend DirectML
Device Intel(R) Arc(TM) Graphics

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 1257
235.1 IPS
 
Image Classification (F16)
100% 1153
215.8 IPS
 
Image Classification (I8)
100% 1038
194.2 IPS
 
Image Segmentation (F32)
100% 1368
22.8 IPS
 
Image Segmentation (F16)
100% 1387
23.2 IPS
 
Image Segmentation (I8)
98% 1186
19.8 IPS
 
Pose Estimation (F32)
100% 12337
14.9 IPS
 
Pose Estimation (F16)
100% 11630
14.1 IPS
 
Pose Estimation (I8)
100% 10196
12.3 IPS
 
Object Detection (F32)
100% 1056
78.9 IPS
 
Object Detection (F16)
100% 1074
80.2 IPS
 
Object Detection (I8)
54% 951
71.0 IPS
 
Face Detection (F32)
100% 1054
12.5 IPS
 
Face Detection (F16)
100% 1066
12.7 IPS
 
Face Detection (I8)
89% 957
11.4 IPS
 
Depth Estimation (F32)
100% 6890
53.4 IPS
 
Depth Estimation (F16)
100% 7115
55.2 IPS
 
Depth Estimation (I8)
95% 5987
46.4 IPS
 
Style Transfer (F32)
100% 36707
48.3 IPS
 
Style Transfer (F16)
100% 37495
49.3 IPS
 
Style Transfer (I8)
98% 28077
36.9 IPS
 
Image Super-Resolution (F32)
100% 5474
195.5 IPS
 
Image Super-Resolution (F16)
100% 5472
195.4 IPS
 
Image Super-Resolution (I8)
99% 3921
140.0 IPS
 
Text Classification (F32)
90% 420
604.0 IPS
 
Text Classification (F16)
90% 472
677.8 IPS
 
Text Classification (I8)
90% 392
563.0 IPS
 
Machine Translation (F32)
100% 1269
23.4 IPS
 
Machine Translation (F16)
100% 1268
23.3 IPS
 
Machine Translation (I8)
70% 754
13.9 IPS
 

Compare

Set Baseline Geekbench 6 now available