Machine Learning Research Intern

About the Role Lightmatter is seeking a PhD Intern to join our Machine Learning team. This internship offers a unique opportunity to contribute to the invention of novel hardware systems utilizing Lightmatter's cutting-edge technology and to model their performance on both training and inference workloads. This is a highly creative and impactful role where you'll be encouraged to think outside the box, pushing the boundaries of ML systems research by leveraging the power of silicon photonics. You'll work closely with our product, solutions architecture, and engineering teams, contributing to: Performance analysis of Lightmatter-enabled systems. The design of high-performance, rack-scale platforms. The generation of cutting-edge ML systems research. Responsibilities Research and stay current on the latest advancements in Machine Learning systems and models, particularly Large Language Models (LLMs). Develop and utilize tools to model the performance of Lightmatter-enabled systems running various training and inference workloads. Collaborate on the design of compelling reference platforms systems powered by Lightmatter technology. Work towards publishing academic papers at top industry conferences or journals, and assist in creating whitepapers, blog posts, and other technical collateral. Minimum Qualifications Currently pursuing a PhD in Computer Science, Electrical Engineering, or a closely related field. Solid academic or research background in Machine Learning systems research, including experience with performance evaluation and modeling. Hands-on experience with machine learning/deep learning methods and knowledge of Large Language Model training and inference frameworks. Familiarity with modern AI systems infrastructure, including compute (e.g., GPUs, TPUs) and networking technologies (e.g., Ethernet, Infiniband) and their various topologies and trade-offs. Demonstrated ability to independently own a research project and drive it from concept to completion within the internship period.   Preferred Attributes Prior academic publications in relevant venues are a strong plus. A demonstrated ability for creative problem-solving and applying first-principles thinking to generate novel ideas in an ambiguous environment. Excellent skills in communicating complex research and technical concepts clearly and effectively. A strong desire to collaborate across research and engineering functions to bridge gaps and accelerate outcomes.