Advanced Technology: AI/ML Research Scientist

Cerebras Systems builds the world's largest AI chip, 56 times larger than GPUs. Our novel wafer-scale architecture provides the AI compute power of dozens of GPUs on a single chip, with the programming simplicity of a single device. This approach allows Cerebras to deliver industry-leading training and inference speeds and empowers machine learning users to effortlessly run large-scale ML applications, without the hassle of managing hundreds of GPUs or TPUs. Cerebras' current customers include top model labs, global enterprises, and cutting-edge AI-native startups. OpenAI recently announced a multi-year partnership with Cerebras , to deploy 750 megawatts of scale, transforming key workloads with ultra high-speed inference. Thanks to the groundbreaking wafer-scale architecture, Cerebras Inference offers the fastest Generative AI inference solution in the world, over 10 times faster than GPU-based hyperscale cloud inference services. This order of magnitude increase in speed is transforming the user experience of AI applications, unlocking real-time iteration and increasing intelligence via additional agentic computation. About The Team Cerebras builds wafer-scale AI processors—single chips delivering tens of PB/s of memory bandwidth and a dataflow architecture that accelerates at a granularity no multi-device system can match. The Advanced Technology Group (ATG) is Cerebras ’ pathfinding organization. We work ahead of product to explore new architectures, demonstrate breakthrough performance on scientific and AI workloads, and shape the technical roadmap for future Cerebras hardware and software. Our work regularly appears at top-tier venues (Supercomputing, SIAM, IEEE, and NeurIPS ) and directly influences the design of next-generation wafer-scale systems. 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 an