Advanced Technology: R&D Engineer - AI/ML, HPC

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. 

We are seeking R&D Engineers to join Cerebras' Advanced Technology Group. You will design and implement workloads that establish new performance benchmarks on wafer-scale hardware, leveraging architectural features that no traditional platform offers. The 
scope ranges from large-scale scientific simulations to emerging AI/ML models, and the work sits at the intersection of algorithm design, compiler co-optimization, and hardware architecture. You will collaborate closely with Cerebras’ ASIC, compiler, kernel, and AI teams as well as external partners at universities and national laboratories. 

• Design and implement challenging scientific computing and AI workloads on Cerebras’ Wafer-Scale Engine, targeting performance results that advance the state of the art. 
• Lead algorithm–hardware co-design efforts with internal R&D teams and external research partners, turning architectural capabilities into measurable application-level advantages. 
• Build analytical performance models that quantify bottlenecks, guide optimization, and inform future chip and compiler design decisions. 
• Contribute to Cerebras’ multi-year technology roadmap by identifying high-impact workloads, proposing architectural experiments, and validating them on silicon. 
• Publish findings and present at top-tier conferences and journals; represent Cerebras in the broader HPC and AI research communities. 

• PhD in Computer Science, Engineering, Applied Mathematics, 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. 
• Deep experience in at least one of the following: computer architecture and accelerator design; parallel, distributed, or high-performance computing; numerical methods and scientific simulation; AI/ML theory and model design at a mathematical level. 
• Strong ability to analytically model and optimize the performance of complex systems and algorithms. 
• Track record of published research or patents in relevant venues. 
• Proficiency in C and Python; comfort working close to hardware. 
• Excellent communication and interpersonal skills: able to present complex technical material to both specialist and cross-functional audiences, and to collaborate effectively in a fast-paced, small-team environment. 

We are hiring across several focus areas. Exceptional depth in one or more of the following is a strong signal: 

• Computational science: researchers who can bring insights from numerical methods and simulation into AI, or couple simulation and learning into joint computational workflows. Depth in hydrodynamics, solid mechanics, electromagnetics, molecular dynamics, or related PDE-based fields. 
• AI/ML foundations: deep understanding of model architecture, optimization methods, and their statistical underpinnings—the ability to design from first principles, not just apply established recipes. 
• Computer architecture: microarchitecture design, computing paradigms at the circuit and datapath level, memory hierarchy design. 
• Performance engineering: roofline modeling, bandwidth analysis, kernel optimization, communication-computation overlap, and compiler-level tuning for novel hardware. 

• Build on a fundamentally different architecture, unconstrained by GPU assumptions. 
• Publish and open-source your research. We present at Supercomputing, SIAM, IEEE, NeurIPS, and beyond. 
• Work on the fastest AI system in the world, with direct access to the hardware your compiler targets. 
• Join at a pivotal moment: Cerebras is pre-IPO with strong commercial traction and rapid growth. 
• Be part of a small, technical team with high autonomy, minimal bureaucracy, and a culture that values depth over hierarchy. 

We are hiring for multiple positions across experience levels. If this work resonates, we encourage you to apply.