Code for paper "Joint Architecture Design and Workload Partitioning for DNN Inference on Industrial IoT Clusters"
-
Updated
Jun 17, 2026 - Python
Code for paper "Joint Architecture Design and Workload Partitioning for DNN Inference on Industrial IoT Clusters"
Trying to shard big embedding tables in multiple devices paying attention to the communication aspects of parallel inference
This dual-MCU repository features a heterogeneous architecture pairing an ESP32-P4 application processor with an ESP32-C5 Wi-Fi 6/BLE hub. Linked over a high-bandwidth 4-bit parallel SDIO interconnect, this design decouples heavy multimedia task execution from the networking stack to completely eliminate application interrupt starvation.
Add a description, image, and links to the parallel-inference topic page so that developers can more easily learn about it.
To associate your repository with the parallel-inference topic, visit your repo's landing page and select "manage topics."