Coosea SN509A vs Coosea SN509A
| Coosea SN509A | Coosea SN509A | Difference | |
|---|---|---|---|
| Single Precision | 445 | 445 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Half Precision | 433 | 433 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Quantized | 872 | 872 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Geekbench AI 1.2.0 | Geekbench AI 1.2.0 | ||
System Information
| Coosea SN509A | Coosea SN509A | |
|---|---|---|
| AI Framework | TensorFlow Lite | TensorFlow Lite |
| AI Backend | CPU | CPU |
| AI Device | ARM ARMv8 | ARM ARMv8 |
| Operating System | Android 14 | Android 14 |
| Model | Coosea SN509A | Coosea SN509A |
| Processor |
ARM ARMv8 @ 2.00 GHz
1 Processor, 8 Cores |
ARM ARMv8 @ 2.00 GHz
1 Processor, 8 Cores |
| Processor ID | ARM implementer 65 architecture 8 variant 4 part 3339 revision 0 | ARM implementer 65 architecture 8 variant 4 part 3339 revision 0 |
| Motherboard | SN509A | SN509A |
| Memory | 5.53 GB | 5.53 GB |
AI Performance
| Coosea SN509A | Coosea SN509A | Difference | |
|---|---|---|---|
| Image Classification (SP) | 211 | 211 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Classification (HP) | 208 | 208 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Classification (Q) | 604 | 604 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Segmentation (SP) | 258 | 258 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Segmentation (HP) | 257 | 257 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Segmentation (Q) | 683 | 683 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Pose Estimation (SP) | 910 | 910 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Pose Estimation (HP) | 903 | 903 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Pose Estimation (Q) | 1568 | 1568 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Object Detection (SP) | 159 | 159 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Object Detection (HP) | 138 | 138 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Object Detection (Q) | 568 | 568 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Face Detection (SP) | 651 | 651 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Face Detection (HP) | 647 | 647 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Face Detection (Q) | 1416 | 1416 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Depth Estimation (SP) | 838 | 838 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Depth Estimation (HP) | 736 | 736 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Depth Estimation (Q) | 1408 | 1408 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Style Transfer (SP) | 1648 | 1648 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Style Transfer (HP) | 1639 | 1639 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Style Transfer (Q) | 3439 | 3439 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Super-Resolution (SP) | 431 | 431 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Super-Resolution (HP) | 424 | 424 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Image Super-Resolution (Q) | 812 | 812 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Text Classification (SP) | 243 | 243 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Text Classification (HP) | 248 | 248 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Text Classification (Q) | 433 | 433 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Machine Translation (SP) | 418 | 418 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Machine Translation (HP) | 426 | 426 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
| Machine Translation (Q) | 288 | 288 | 100.0% |
| Coosea SN509A | |||
| Coosea SN509A | |||
