Research Intern, Physical AI and GPU (PhD)

About Marvell Marvell’s semiconductor solutions are the essential building blocks of the data infrastructure that connects our world. Across enterprise, cloud and AI, and carrier architectures, our innovative technology is enabling new possibilities. 

At Marvell, you can affect the arc of individual lives, lift the trajectory of entire industries, and fuel the transformative potential of tomorrow. For those looking to make their mark on purposeful and enduring innovation, above and beyond fleeting trends, Marvell is a place to thrive, learn, and lead. 

Your Team, Your Impact Marvell is a global leader in high-performance silicon and AI infrastructure, powering the world's most advanced datacenters, cloud platforms, 5G networks, and AI/ML workloads. Our innovations drive next-generation computing, networking, and storage — enabling breakthrough performance, scalability, and efficiency for the most demanding applications.

Why This Internship

• Engage with open research problems at the frontier of AI systems — challenges that require access to real silicon, large-scale infrastructure, and production data
• Contribute to work intended for publication, with support for presenting results both internally and externally
• Translate research outcomes into real products and next-generation silicon, extending impact beyond publication
• Collaborate with and receive mentorship from senior architects and researchers who will help advance your ideas
• Gain access to advanced infrastructure, tools, and pre-release hardware rarely available in academic settings
• Pursue original, exploratory research directions — independent thinking is actively encouraged What You Can Expect Focus: Build the platforms and infrastructure that Physical AI runs on — the intelligence, networking, and silicon that let autonomous systems perceive, reason, and act in the real world, on tightly constrained compute.

Key Responsibilities Prototype and prove what next-generation Physical AI platforms can do Write and optimize CUDA to extract maximum performance from limited memory and low-end GPUs Build the compute, networking, and security platforms that autonomous systems run on Engineer agentic capabilities — systems that plan, decide, and act on their own Design and train ML models and SLMs for perception, reasoning, and action on real-world input (video, audio, sensors)

What We're Looking For Eligibility: Enrolled in a PhD, MS by Research, or MTech by Research program in Computer Science, Computer Engineering, electronics or Electrical Engineering, Robotics or a related field.

Requirements Strong GPU and CUDA programming skills — kernel optimization, memory management, and performance tuning under tight resource constraints Strong fundamentals in machine learning, computer architecture, and systems Experience with robotic platforms or real-world AI applications (perception, sensors, or real-time control) Programming in C/C++ and Python Hands-on with ML modeling, training, and fine-tuning (PyTorch or TensorFlow)

Preferred: Exposure to model optimization and quantization for constrained hardware, robotics frameworks (e.g., ROS), on-device inference, agentic systems, or next-generation AI silicon.

Additional Compensation and Benefit Elements With competitive compensation and great benefits, you will enjoy our workstyle within an environment of shared collaboration, transparency, and inclusivity. We’re dedicated to giving our people the tools and resources they need to succeed in doing work that matters, and to grow and develop with us. For additional information on what it’s like to work at Marvell, visit our Careers page.

All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, national origin, sexual orientation, gender identity, disability or protected veteran status.

Interview Integrity  To support fair and authentic hiring practices, candidates are not permitted to use AI tools (such as transcription apps, real-time answer generators like ChatGPT or Copilot, or automated note-taking bots) during interviews.

These tools must not be used to record, assist with, or enhance responses in any way. Our interviews are designed to evaluate your individual experience, thought process, and communication skills in real time. Use of AI tools without prior instruction from the interviewer will result in disqualification from the hiring process.

This position may require access to technology and/or software subject to U.S. export control laws and regulations, including the Export Administration Regulations (EAR). As such, applicants must be eligible to access export-controlled information as defined under applicable law. Marvell may be required to obtain export licensing approval from the U.S. Department of Commerce and/or the U.S. Department of State. Except for U.S. citizens, lawful permanent residents, or protected individuals as defined by 8 U.S.C. 1324b(a)(3), all applicants may be subject to an export license review process prior to employment.

#LI-TT1