Radar Systems GPU Software engineer
Overview:
In this role, you design and optimize GPU-accelerated radar signal processing for automotive ADAS and autonomous driving. You will work within a cross-functional team to deliver real-time radar algorithms on GPU platforms and ensure robust performance in vehicle environments. You'll diagnose bottlenecks, test integrations with ADAS/autonomous systems, and document complex solutions for broad use. Aptiv offers a mission-driven environment focused on safer, software-defined mobility and scalable radar technologies. This is a hands-on, impact-focused opportunity to shape next-generation automotive sensing capabilities. Compensation / Benefits Private health care effective day 1Life and accident insurance Paid Time OffRelocation assistance may be available Learning and development opportunities Tuition Reimbursement
Responsibilities:
- Design, implement, and optimize radar signal processing algorithms for GPU-based automotive platforms
- Enhance performance using GPU parallelism and memory management to achieve real-time, low-latency radar processing
- Integrate radar algorithms with ADAS and autonomous driving systems and validate them with cross-functional teams
- Debug and troubleshoot performance issues and bugs on GPU platforms
- Collaborate with radar system engineers, hardware teams, and software developers to align requirements
- Document technical designs, code, and usage guides for cross-team clarity
Key requirements:
- Master's degree in Computer Science or Electrical Engineering 3+ years of GPU programming and radar signal processing experience in automotive or related industriesCUDA expertise with a focus on real-time processing
- Strong knowledge of radar algorithms (FFT, Doppler, CFAR) and radar sensor data
- Familiarity with ADAS standards (ISO 26262, AUTOSAR) and radar system requirements
- Proficiency in C/C++, Python, and MATLABExperience with GPU debugging/profiling tools such as Nsight, PerfHUDcollaborative team playerstrong communicationproblem solvingCUDAradar signal processingFFT