Ph.D. Intern - Silicon Photonics & Optical Communications

Marvell Technology•Irvine, CA

2d•$37 - $73

About The Position

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 The AI era runs on light. Every rack of XPUs training a frontier model, every hyperscaler connecting data centers across a metro region, every 1.6T optical module pushing data across a fiber at speeds that didn't exist three years ago — all of it depends on silicon photonics and optical DSP technology that Marvell designs. Marvell's optical team is one of the few places in the world where silicon photonics, coherent DSP, PAM4 signal processing, and co-packaged optics integration are all happening under one roof, at production scale, for the world's largest AI infrastructure operators. The work spans every layer of the AI interconnect hierarchy. At the scale-up layer — connecting XPUs within a rack or cluster to behave as a single logical compute unit — the team is developing silicon photonics light engines for Co-Packaged Optics and Near-Packaged Optics platforms that eliminate the reach and power limitations of copper at the shortest distances. At the scale-out layer — connecting clusters of thousands of XPUs across rows and data center floors — the Ara 3nm PAM4 DSP platform drives 1.6T optical transceivers that set the standard for bandwidth density and power-per-bit efficiency across AI fabric networks. And at the scale-across layer — connecting geographically distributed data centers across metro and regional distances — the Orion 800G coherent DSP powers the COLORZ 800, the industry's first 800ZR/ZR+ pluggable coherent transceiver, carrying 800 Gbps up to 1000km over a single wavelength. These are not incremental improvements. They are fundamental rearchitectings of how data moves inside and between the world's most advanced AI infrastructure. Marvell's Ph.D. Intern Program places doctoral candidates directly inside these active design and research efforts. Projects are selected because they sit at the intersection of Marvell's most pressing optical engineering challenges and the photonics, signal processing, and device physics that define doctoral research in this field. The work is the applied dimension of the academic research a Ph.D. candidate is already pursuing — conducted on production silicon, against real system specifications, for hyperscale customers who are counting on it. What you will take away is something no lab or simulation environment can replicate: the experience of seeing your research become working optical silicon, deployed at scale inside the world's most advanced AI infrastructure.

Requirements

Currently enrolled in a Ph.D. program in Electrical Engineering, Applied Physics, or a related field, with a research focus in photonics, optical communications, or high-speed signal processing
Demonstrate research experience in one or more of the following: silicon photonics device design and simulation, optical transceiver systems, coherent or direct-detect optical communications, or DSP algorithm development for high-speed links
Apply strong fundamentals in photonics and electromagnetics — waveguide theory, modulator physics, photodetector design, or optical link analysis
Use simulation and modeling tools relevant to your research area (FDTD, MODE, Lumerical, MATLAB, or equivalent) with confidence and independence
Communicate research findings and design tradeoffs clearly — you will present your work to the engineering team and defend your approach

Nice To Haves

Experience with PAM4 or coherent modulation formats, FEC, or DSP equalization techniques for optical links
Familiarity with silicon photonics fabrication processes and design-for-manufacturing considerations in advanced CMOS-compatible platforms
Exposure to co-packaged optics, near-packaged optics, or optical engine integration for high-density interconnects
Lab experience with high-speed optical or electrical test and measurement equipment
Familiarity with analog IC design or mixed-signal circuits as they relate to optical transceiver front-ends (TIAs, laser drivers, modulators) is a strong plus

Responsibilities

Design, simulate, and characterize silicon photonics components and subsystems — including modulators, photodetectors, waveguides, and optical engines — for integration into PAM4 and coherent transceiver platforms
Develop and validate DSP algorithms for high-speed optical links, including equalization, FEC, and signal integrity techniques for PAM4 and coherent modulation formats
Work on optical system modeling and link budget analysis for data center interconnect, co-packaged optics, and near-packaged optics applications
Collaborate with analog IC design, packaging, and systems engineering teams to bring photonic and electronic components together into integrated transceiver solutions
Characterize and validate designs in the lab — working with high-speed test equipment, optical measurement tools, and silicon bring-up setups
Participate in a structured design curriculum alongside your Ph.D. intern peers and present your research findings and design results to the engineering team