Research and implementation of on-device LLM inference on NXP i.MX95, exploring Neutron NPU acceleration, ONNX Runtime, INT4 quantization, hybrid CPU/NPU execution, and performance optimization.
arm64 embedded-ai embedde model-quantization embedded-ai-systems embedded-llm neutron-npu nxp-imx95 embedded-nxp
-
Updated
Nov 13, 2026 - Python