About The Role Most AI research today is shaped by the constraints of existing hardware. This role starts from the other direction: what would you build if the architecture let you rethink the fundamentals? You will design and develop AI models and training methodologies on wafer-scale hardware, working at the level of optimization theory, model architecture, and statistical foundations rather than assembling existing components. The ATG sits at the intersection of AI, computational science, and computer architecture, and your work will draw on all three. You will collaborate closely with Cerebras’ ASIC, compiler, kernel, and AI teams as well as external partners at universities and national laboratories. What You Will Do Design AI models and training methods from first principles, leveraging architectural properties of wafer-scale hardware that are unavailable on conventional platforms. Investigate how techniques from computational science—numerical methods, PDE solvers, simulation—can inform and advance AI model design, and explore hybrid workflows that couple simulation and learning. Develop a deep understanding of the hardware substrate and use it to guide algorithmic choices: model structure, optimization strategy, memory access patterns, numerical precision. Publish findings and present at top-tier venues (NeurIPS, ICML, ICLR, etc.); represent Cerebras in the broader AI/ML research community. Inform the design of future Cerebras hardware and software by identifying the computational patterns that matter most for next-generation AI workloads. What We Are Looking For PhD in Machine Learning, Computer Science, Applied Mathematics, Statistics, Physics, or a related quantitative field preferred ; exceptional candidates without a graduate degree who demonstrate equivalent depth through published research, significant open-source contributions, or a strong industry track record are encouraged to apply. Mathematical maturity: comfort with the theory behind gradient methods, loss landscapes, generalization, and the relationship between model structure and data statistics. Track record of published research at top-tier AI or computational science venues. Proficiency in Python and PyTorch; comfort with C or other low-level languages is a strong signal. Excellent communication and interpersonal skills: able to present complex technical material to both ML and systems audiences, and to collaborate effectively in a fast-paced, small-team environment. Why This Opportunity Is Exciting You will have direct access to hardware that changes what’s algorithmically possible. Tens of PB/s of memory bandwidth and fine-grained dataflow execution open design spaces that don’t exist on GPU clusters. You will work alongside researchers in computational science, computer architecture, and performance engineering. The synthesis across these fields is central to ATG’s approach. Your research will influence silicon - ATG’s findings directly shape the design of future Cerebras chips and systems. Equal Opportunity Statement Cerebras Systems is committed to creating an equal and diverse environment and is proud to be an equal opportunity employer. We celebrate different backgrounds, perspectives, and skills. We believe inclusive teams build better products and companies. We try every day to build a work environment that empowers people to do their best work through continuous learning, growth and support of those around them. #J-18808-Ljbffr