Samsung Galaxy S23 vs GMKtec NucBoxG9
| Samsung Galaxy S23 | GMKtec NucBoxG9 | Difference | |
|---|---|---|---|
| Single Precision | 1706 | 601 | 283.9% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Half Precision | 1764 | 612 | 288.2% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Quantized | 2603 | 430 | 605.3% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Geekbench AI 1.1.0 | Geekbench AI 1.3.0 | ||
System Information
| Samsung Galaxy S23 | GMKtec NucBoxG9 | |
|---|---|---|
| AI Framework | TensorFlow Lite | TensorFlow Lite |
| AI Backend | CPU | CPU |
| AI Device | Qualcomm Snapdragon 8 Gen 2 | Intel(R) N150 |
| Operating System | Android 14 | Ubuntu 24.04.2 LTS |
| Model | Samsung Galaxy S23 | GMKtec NucBoxG9 |
| Processor |
ARM ARMv8 @ 2.02 GHz
1 Processor, 8 Cores |
Intel N150 @ 3.60 GHz
1 Processor, 4 Cores |
| Processor ID | ARM implementer 65 architecture 8 variant 1 part 3406 revision 0 | GenuineIntel Family 6 Model 190 Stepping 0 |
| Motherboard | kalama | GMKtec GMKtec |
| BIOS | American Megatrends International, LLC. 5.27 | |
| Memory | 6.89 GB | 11.42 GB |
AI Performance
| Samsung Galaxy S23 | GMKtec NucBoxG9 | Difference | |
|---|---|---|---|
| Image Classification (SP) | 1439 | 406 | 354.4% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Classification (HP) | 1754 | 411 | 426.8% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Classification (Q) | 2649 | 275 | 963.3% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Segmentation (SP) | 1570 | 517 | 303.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Segmentation (HP) | 1718 | 515 | 333.6% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Segmentation (Q) | 2805 | 332 | 844.9% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Pose Estimation (SP) | 2673 | 632 | 422.9% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Pose Estimation (HP) | 2789 | 649 | 429.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Pose Estimation (Q) | 5574 | 611 | 912.3% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Object Detection (SP) | 1305 | 369 | 353.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Object Detection (HP) | 1299 | 378 | 343.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Object Detection (Q) | 1748 | 286 | 611.2% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Face Detection (SP) | 2540 | 771 | 329.4% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Face Detection (HP) | 2541 | 870 | 292.1% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Face Detection (Q) | 3983 | 719 | 554.0% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Depth Estimation (SP) | 2511 | 724 | 346.8% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Depth Estimation (HP) | 2586 | 747 | 346.2% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Depth Estimation (Q) | 4170 | 549 | 759.6% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Style Transfer (SP) | 6074 | 1256 | 483.6% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Style Transfer (HP) | 6114 | 1264 | 483.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Style Transfer (Q) | 10563 | 1454 | 726.5% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Super-Resolution (SP) | 1239 | 362 | 342.3% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Super-Resolution (HP) | 1239 | 338 | 366.6% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Image Super-Resolution (Q) | 2639 | 357 | 739.2% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Text Classification (SP) | 494 | 612 | 80.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Text Classification (HP) | 482 | 614 | 78.5% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Text Classification (Q) | 829 | 266 | 311.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Machine Translation (SP) | 1120 | 819 | 136.8% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Machine Translation (HP) | 1114 | 833 | 133.7% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
| Machine Translation (Q) | 515 | 253 | 203.6% |
| Samsung Galaxy S23 | |||
| GMKtec NucBoxG9 | |||
