Geekbench ML Score

6377
ONNX DirectML Inference Score
Geekbench ML 0.6.0 for Windows x86 (64-bit)

Result Information

Upload Date December 17 2024 10:12 PM
Views 4

System Information

System Information
Operating System Microsoft Windows 10 Pro (64-bit)
Model Apple Inc. MacPro5,1
Motherboard Apple Inc. Mac-F221BEC8
Power PlanBalanced
CPU Information
Name Intel Xeon E5620
Topology 2 Processors, 8 Cores, 16 Threads
Identifier GenuineIntel Family 6 Model 44 Stepping 2
Base Frequency 2.53 GHz
Cluster 18 Cores
L1 Instruction Cache32.0 KB x 4
L1 Data Cache32.0 KB x 4
L2 Cache256 KB x 4
L3 Cache12.0 MB x 1
Memory Information
Size 32.00 GB
Inference Information
Framework ONNX
Backend DirectML
Device AMD Radeon VII

Inference Performance

Workload Accuracy Score
Image Classification (F32)
100% 3476
650.4 IPS
 
Image Classification (F16)
100% 3441
644.0 IPS
 
Image Classification (I8)
99% 2982
558.0 IPS
 
Image Segmentation (F32)
100% 2862
47.8 IPS
 
Image Segmentation (F16)
100% 2864
47.8 IPS
 
Image Segmentation (I8)
98% 2715
45.3 IPS
 
Pose Estimation (F32)
100% 87992
106.6 IPS
 
Pose Estimation (F16)
100% 84913
102.8 IPS
 
Pose Estimation (I8)
100% 73032
88.4 IPS
 
Object Detection (F32)
100% 1375
102.6 IPS
 
Object Detection (F16)
100% 1353
101.0 IPS
 
Object Detection (I8)
61% 1257
93.9 IPS
 
Face Detection (F32)
100% 8178
97.2 IPS
 
Face Detection (F16)
100% 7835
93.2 IPS
 
Face Detection (I8)
88% 7429
88.3 IPS
 
Depth Estimation (F32)
100% 21400
166.0 IPS
 
Depth Estimation (F16)
100% 21295
165.1 IPS
 
Depth Estimation (I8)
95% 18184
141.0 IPS
 
Style Transfer (F32)
100% 105602
138.9 IPS
 
Style Transfer (F16)
100% 102506
134.8 IPS
 
Style Transfer (I8)
98% 89808
118.1 IPS
 
Image Super-Resolution (F32)
100% 9876
352.7 IPS
 
Image Super-Resolution (F16)
100% 10229
365.3 IPS
 
Image Super-Resolution (I8)
99% 8063
287.9 IPS
 
Text Classification (F32)
100% 1077
1.55 KIPS
 
Text Classification (F16)
100% 1065
1.53 KIPS
 
Text Classification (I8)
98% 665
955.2 IPS
 
Machine Translation (F32)
100% 907
16.7 IPS
 
Machine Translation (F16)
100% 911
16.8 IPS
 
Machine Translation (I8)
69% 693
12.7 IPS
 

Compare

Set Baseline Geekbench 6 now available