Kernel Engineer - New Grad

About the role

Cerebras Systems builds the world's largest AI chip, 56 times larger than GPUs. This architecture allows Cerebras to deliver industry-leading training and inference speeds; 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.

Cerebras works with the leading 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.

About the Role

As a Kernel Engineer at Cerebras, you will develop high-performance software at the intersection of hardware and software for cutting-edge artificial intelligence and high-performance computing workloads.

You will help implement, optimize, and validate machine learning and linear algebra operations for the Cerebras Wafer-Scale Engine, our custom massively parallel processor architecture. Working alongside experienced kernel, compiler, performance, and hardware engineers, you will learn how algorithms are mapped to specialized hardware and contribute to software that maximizes compute utilization and system performance.

You will be part of a team responsible for the design, development, performance tuning, and validation of foundational ML and HPC kernels. This is an excellent opportunity for a new graduate who is interested in computer architecture, parallel programming, low-level software, and machine learning systems.

Responsibilities

• Help design and implement machine learning and linear algebra kernels for the Cerebras Wafer-Scale Engine.
• Develop and debug high-performance kernel routines using low-level programming techniques and the Cerebras Software Language, a custom C-like language.
• Apply parallel programming algorithms to map computational workloads efficiently onto the Cerebras architecture.
• Use mathematical analysis, performance data, and profiling tools to evaluate kernel behavior and inform design decisions.
• Identify and investigate correctness, performance, and hardware utilization issues.
• Develop unit tests and system-level validation methodologies to verify the functionality and performance of kernel libraries.
• Collaborate with kernel, compiler, performance, and hardware engineers to improve software and system performance.
• Study emerging machine learning workloads and contribute to the evolution of the kernel library.
• Participate in code reviews, technical discussions, and software development processes.
• Build an understanding of the Cerebras architecture, instruction set, memory system, and communication model.

Minimum Skills & Qualifications

• Bachelor’s, Master’s, or PhD in Computer Science, Computer Engineering, Electrical Engineering, Mathematics, or a related field.
• Strong programming fundamentals in C++ and familiarity with Python.
• Understanding of foundational computer architecture concepts such as processors, memory hierarchies, instruction execution, or data movement.
• Knowledge of data structures, algorithms, and software development fundamentals.
• Experience debugging software through coursework, internships, research, co-op placements, or technical projects.
• Strong analytical and problem-solving skills.
• Interest in low-level software, parallel computing, performance optimization, or hardware/software co-design.
• Ability to learn unfamiliar systems and collaborate effectively within a technical team.

Preferred Skills & Qualifications

• Research, internships, or projects involving kernel development, compilers, computer architecture, HPC, or systems programming.
• Familiarity with parallel algorithms, multithreaded programming, or distributed memory systems.
• Exposure to programming accelerators such as GPUs, FPGAs, or other specialized processors.
• Experience with low-level programming, assembly language, CUDA, OpenCL, or a domain-specific language.
• Familiarity with machine learning concepts, neural networks, or frameworks such as PyTorch or TensorFlow.
• Exposure to numerical computing, linear algebra, or HPC kernels.
• Experience using profiling, benchmarking, or performance analysis tools.
• Familiarity with library or API development practices.