Table 13.
Machine learning accelerators.
| Accelerator | Processor Type | RAM | Flash Memory | Power | Application | Maker |
|---|---|---|---|---|---|---|
| ARM Ethos NPU | ARM ML processor | 1 GB | - | 4 TOPs/W | Image processing, voice recognition | ARM |
| BeagleBone AI | Cortex-A15, Sitara AM5729 | 1 GB | 16 GB | 500 mW | Computer vision | Texas Instruments |
| Google Coral Dev Board | GPU, TPU, CPU | 1 GB | 8 GB | (6–10) W | Image Processing | |
| Intel Movidus NCS | VPU | 1 GB | 4 GB | 500 mW | Classification, computer vision | Intel |
| Mustang-F100-A10 | Intel Arria 10 GX1150 FPGA | 8 GB | - | 40 W | Computer vision | Intel |
| NVIDIA Jetson NANO | GPU, CPU | 4 GB 64 it LPDDR4 25.6 GB/s | 16 GB eMMC, 5.1 Flash | (5–10) W | Computer vision, audio processing | NVIDIA |