An article written by Frederik Bode about the first benchmark of Google’s EdgeTPU Dev Board is in. Read about comparison is made against Intel’s (first generation) Movidius Neural Compute Stick, and Google is the clear winner regarding inference time.
The Edge TPU is a small ASIC designed by Google that provides high performance ML inferencing for low-power devices. For example, it can execute state-of-the-art mobile vision models such as MobileNet V2 at 100+ fps, in a power efficient manner.
At the moment Edge TPU supports TensorFlow Lite only.
The article then describes:
- Benchmark setup including parents and children
- The data source
- Usability and plug-and-play
- Advice around troubleshooting
In conclusion, EdgeTPU far outperforms the first generation Movidius Compute Stick, and confidently awaits its next contender, Intel’s Movidius Neural Compute Stick 2. The speedup generated by the EdgeTPU Dev Board is ~13x. Great for any Raspberry Pi enthusiasts.
[Read More]