Week #7 — Hardware Changes
Sachin C -
Hello, and welcome back!
This week, I finished the model that I intend on testing the controllers on. I finalized the dataset, and trained the final version of the model using the latest architecture.
Additionally, I made some changes to the hardware I intend on testing.
The Texas Instruments AM69 board I detailed in a previous post is no longer being used in the experiment. This is because of the price of the board, which starts at $400, and the shipping, which would not arrive before May. I intend on using a different chipset — the NXP i.MX 8M Plus. Instead of the $400 price tag, it ranges between $90 and $120, and can ship before my project is due.
The NXP board is extremely similar to the AM69 for a few reasons:
-
Quad-core ARM Cortex-A53 CPU (similar performance class)
-
Neural Processing Unit (NPU) for AI acceleration (2.3 TOPS)
-
Supports TensorFlow Lite, ONNX, and PyTorch models (TensorFlow is what I used for my model.)
Thanks to my incredible site placement, I was lucky enough to be able to use their Raspberry Pi 5, which is the upgraded version of the Pi 4 I intended on using! The Arduino Portenta H7 and the USB-C power meter I need to measure the voltage also arrived in the mail yesterday, and I will pick up the ESP32 board today. Next week, I intend on porting the model to the various operating systems each controller uses, and finally running the experiment!
I hope to see you next week. Thank you!
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.