Distributed 56M-parameter LLM inference across 3 ESP32-S3 boards via ESP-NOW , Split-PLE + KV cache, fully offline.
embedded optimization esp32 transformers embedded-systems transformer kv embedded-c edge-ai ondeviceai ondevicemachinelearning quantilization esp32s3 esp32-s3 llm llm-inference edge-ai-models edge-ai-engineering ondevice-ai
-
Updated
Jul 30, 2026 - C