Manager, Kernel Software
CerebrasJob Title
Manager, Kernel Software
Role Summary
Lead a kernel engineering team to design, implement, and optimize high-performance ML and HPC kernels for a custom, massively parallel processor. Work at the intersection of hardware and software to maximize compute utilization and improve training and inference efficiency on Cerebras systems.
Experience Level
Senior — managerial role requiring proven technical leadership; specific years of experience not stated.
Responsibilities
Own the design, implementation, validation, and performance tuning of low-level kernel libraries and related tooling to exploit Cerebras hardware.
- Define design specifications for new ML and linear algebra kernels and map algorithms to the Cerebras WSE system.
- Implement and optimize kernel routines in low-level assembly and a custom C-like language (CSL).
- Develop and maintain a library of parallel and distributed algorithms to maximize compute utilization.
- Use mathematical models and performance analysis to guide design decisions and optimization efforts.
- Integrate unit and system testing to verify correctness and performance of kernel libraries.
- Collaborate with chip and system architects to influence instruction sets, microarchitecture, and IO for next-generation systems.
- Monitor emerging ML trends and evolve kernel architecture to address new computational patterns.
- Mentor engineers, set technical direction, and drive execution within the team.
Requirements
Must-have technical skills, leadership experience, and practical abilities required for the role.
- Proven experience leading technical teams and mentoring engineers.
- Strong understanding of hardware architecture concepts and willingness to work on new system architectures.
- Proficiency in C++ and Python and experience with low-level systems programming.
- Familiarity with library/API development best practices and performance optimization techniques.
- Excellent debugging skills across complex, layered software stacks.
- Experience with performance measurement, modeling, and tuning of compute kernels.
- Ability to collaborate with architects and cross-functional teams to align software and hardware design.
Education Requirements
Bachelor’s, Master’s, PhD, or foreign equivalent in Computer Science, Computer Engineering, Mathematics, or a related field.
About the Company
Company: Cerebras
Headquarters: Sunnyvale, CA, USA
Developer of wafer-scale AI accelerators, Cerebras designs the Wafer Scale Engine (WSE)—one of the world’s largest AI chips—to deliver high-speed training and inference solutions for model labs, enterprises, and AI-native startups.
