Senior Quantization Engineer - Edge AI at NXP Semiconductors

Hyderabad, Telangana, India -

Full Time

Start Date

Immediate

Expiry Date

29 Aug, 26

Salary

0.0

Posted On

31 May, 26

Experience

5 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Model Quantization, PyTorch, ONNX, Python, C++, CNN, LLM, VLM, Speculative Decoding, Pruning, Embedded Systems, NPU, MLIR, TVM, Deep Learning, Machine Learning

Industry

Semiconductor Manufacturing

Description

Senior Quantization Engineer - Edge AI Model Optimization We at NXP have an environment that fosters innovation. Our team has technology experts who understand the big picture and mentors who coach passionate professionals to work on the most exciting challenges. We share responsibilities in everything we do, where every point of view is valued. Join us! Job Summary We are seeking a highly skilled Edge AI Engineer/Scientist with a strong theoretical foundation in AI and solid software engineering expertise to contribute to our Edge AI Model Optimization program. While the primary focus of this role is on model quantization, the scope also includes complementary optimization strategies such as speculative decoding, pruning, and other methods for ensuring highly efficient on-device deployment. You will work at the forefront of innovation, bridging the gap between research and practice, focusing on CNNs, Large Language Model (LLM) and Vision Language Model (VLM) optimization to bring advanced capabilities to NXP’s Ara2 family of NPUs, directly supporting the future of on‑device intelligence. If you want to the future of efficient on‑device AI, this is the place to be. Job Responsibilities Research: Actively survey the latest research (NeurIPS, ICLR, CVPR) on model optimization/compression, focusing particularly on neural network quantization, but also including other techniques like speculative decoding, pruning, etc. Prototyping: Develop and adapt state-of-the-art methods to NXP’s hardware constraints, building POCs to showcase the effectiveness of these techniques on NXP HW. Production Implementation: Translate research prototypes into robust, optimized production code (C++/Python), ensuring strict memory and compute efficiency standards. Systems Integration: Document algorithmic tradeoffs, derive deployment recipes, and mentor the engineering team on numerical methods and optimization. Cross-Functional Leadership: Act as the technical bridge between AI Research, Hardware Engineering and other teams, providing quantified guidance on how choices impact model accuracy and performance. IP Generation: Contribute to NXP’s intellectual property portfolio through patents and technical publications. Job Qualifications Required Background Education: MSc or Ph.D (is a plus) in Computer Science, Electrical Engineering, or Mathematics with a focus on Machine Learning or Deep Learning. AI Expertise: Proven practical experience in AI/ML with a deep understanding of CNN architectures and Generative AI (Transformers, LLMs, VLMs, etc.). Technical Stack: Strong hands-on experience with PyTorch, ONNX, and model conversion/optimization pipelines. Software Engineering: Proficient in Python and C++ and best development practices. Embedded Mindset: Familiarity with the constraints of embedded systems (latency, power, memory bandwidth) and how code interacts with underlying hardware. Preferred Advanced AI: Experience with state-of-the-art quantization techniques for discriminative and generative AI (e.g., GPTQ, SpinQuant, etc). Hardware Acceleration: Experience with NPUs, device-level profiling, and diagnosing memory bottlenecks. Kernel Development: Experience with custom kernel development is a plus. Compilers: Knowledge of MLIR or TVM is a significant plus. We at NXP have an environment that fosters innovation. Our team has technology experts who understand the big picture and mentors who coach passionate professionals to work on the most exciting challenges. We share responsibilities in everything we do, where every point of view is valued. Join us! Job Summary We are seeking a highly skilled Edge AI Engineer/Scientist with a strong theoretical foundation in AI and solid software engineering expertise to contribute to our Edge AI Model Optimization program. While the primary focus of this role is on model quantization, the scope also includes complementary optimization strategies such as speculative decoding, pruning, and other methods for ensuring highly efficient on-device deployment. You will work at the forefront of innovation, bridging the gap between research and practice, focusing on CNNs, Large Language Model (LLM) and Vision Language Model (VLM) optimization to bring advanced capabilities to NXP’s Ara2 family of NPUs, directly supporting the future of on‑device intelligence. If you want to the future of efficient on‑device AI, this is the place to be. Job Responsibilities Research: Actively survey the latest research (NeurIPS, ICLR, CVPR) on model optimization/compression, focusing particularly on neural network quantization, but also including other techniques like speculative decoding, pruning, etc. Prototyping: Develop and adapt state-of-the-art methods to NXP’s hardware constraints, building POCs to showcase the effectiveness of these techniques on NXP HW. Production Implementation: Translate research prototypes into robust, optimized production code (C++/Python), ensuring strict memory and compute efficiency standards. Systems Integration: Document algorithmic tradeoffs, derive deployment recipes, and mentor the engineering team on numerical methods and optimization. Cross-Functional Leadership: Act as the technical bridge between AI Research, Hardware Engineering and other teams, providing quantified guidance on how choices impact model accuracy and performance. IP Generation: Contribute to NXP’s intellectual property portfolio through patents and technical publications. Job Qualifications Required Background Education: MSc or Ph.D (is a plus) in Computer Science, Electrical Engineering, or Mathematics with a focus on Machine Learning or Deep Learning. AI Expertise: Proven practical experience in AI/ML with a deep understanding of CNN architectures and Generative AI (Transformers, LLMs, VLMs, etc.). Technical Stack: Strong hands-on experience with PyTorch, ONNX, and model conversion/optimization pipelines. Software Engineering: Proficient in Python and C++ and best development practices. Embedded Mindset: Familiarity with the constraints of embedded systems (latency, power, memory bandwidth) and how code interacts with underlying hardware. Preferred Advanced AI: Experience with state-of-the-art quantization techniques for discriminative and generative AI (e.g., GPTQ, SpinQuant, etc). Hardware Acceleration: Experience with NPUs, device-level profiling, and diagnosing memory bottlenecks. Kernel Development: Experience with custom kernel development is a plus. Compilers: Knowledge of MLIR or TVM is a significant plus. More information about NXP in India... #LI-29f4 NXP Semiconductors N.V. (NASDAQ: NXPI) enables a smarter, safer, and more sustainable world through innovation. As the world leader in secure connectivity solutions for embedded applications, NXP is pushing boundaries in the automotive, industrial & IoT, mobile, and communication infrastructure markets. For more information, visit www.nxp.com Bright Minds. Bright Futures. We believe that a key component to growing our business is to develop our people. To enable you to grow your career at NXP, we offer online and offline learning opportunities to help you develop some of your core and professional skills. Commitment At NXP. We recognize NXP is a powerful change agent as we continue to deliver innovative solutions that advance a more sustainable future. We remain steadfast in our commitment to sustainability and making measurable year-on-year progress. Also, we aim to create an inclusive work environment and we will not tolerate racism, discrimination or harassment of any kind. We have programs in place focused on diversity, inclusion and equality. Thank you for considering a career at NXP. To help you prepare for the different steps in our hiring process, see the following useful advice and tips. Are you already an NXP employee? Do not apply here. Instead, you must apply via our internal career page. Thank you for your interest in supporting our recruitment efforts. Please note that NXP operates under a strict Preferred Supplier List (PSL) for all recruitment activities. Any candidate profiles or resume submitted without a prior written agreement or explicit request from our Talent Acquisition team will be considered unsolicited. Such submissions will be deemed free of any obligations, and no fees will be paid by NXP or any of its affiliates, subsidiaries, or divisions - regardless of whether the candidate is hired, either coincidentally or otherwise. Thank you for your understanding.

Responsibilities

Research and implement state-of-the-art model optimization and quantization techniques for CNNs, LLMs, and VLMs. Translate research prototypes into production-ready C++/Python code for deployment on NXP's Ara2 family of NPUs.