NPU Accelerator Software Engineer

About the role

You will join the eIQ AI Runtime team at NXP, working closely with an international, cross‑functional group of engineers to develop low‑level software that enables NPU accelerators on automotive‑grade MCUs and MPUs.

Key responsibilities

• Develop and integrate low‑level software for NPU accelerators on automotive MCUs/MPUs.
• Integrate NPU functionality into existing BSPs and complex automotive software stacks, ensuring real‑time and safety compliance.
• Work within AUTOSAR and/or embedded Linux environments to support AI/ML workloads.
• Collaborate with hardware, AI/ML, and systems engineering teams to define and optimise accelerator usage.
• Debug, validate and tune performance on target automotive hardware.
• Ensure compliance with ISO 26262, MISRA C/C++ and other automotive safety standards.

Required profile

• Bachelor’s or Master’s degree in Computer Engineering, Electrical Engineering or related field.
• 3–5 years of embedded software development experience using Embedded C/C++.
• Experience with NPU or AI accelerator integration in embedded systems.
• Proficiency in bare‑metal/RTOS and embedded Linux (Yocto, Buildroot).
• Understanding of hardware‑software interfaces, memory‑mapped I/O and DMA.

Required skills

• Embedded C/C++ development.
• RTOS (AUTOSAR) and embedded Linux environments.
• NPU/AI accelerator integration.
• Yocto, Buildroot, bare‑metal programming.
• ISO 26262, MISRA C/C++ compliance.
• Git, CI/CD pipelines, Agile methodologies.
• Familiarity with TensorFlow Lite, ONNX Runtime (optional).
• DSP/HiFi5 firmware and driver development (optional).

What we offer

• Permanent full‑time contract with competitive salary, bonus plan and lunch vouchers.
• Opportunity to purchase company shares at a 15% discount.
• Flexible working hours and work‑from‑home policy.
• Professional development through online and offline learning.
• Modern office in Bucharest with relaxation areas, library, cafeteria and easy public‑transport access.