Next-Gen AI Performance Engineer — WSE Architect

Cerebras Systems, Inc.; Sunnyvale (CA); On-site; USD 150,000 - 190,000; Full time

Job summary

Cerebras Systems, Inc. in Sunnyvale, CA, seeks a performance engineer to optimize AI workloads on the Wafer Scale Engine. You will work across hardware and software to identify bottlenecks and improve efficiency.

Responsibilities include bringing up new WSE generations, building performance models, and debugging kernel code to elevate inference speed and throughput. Strong architecture background and C++/Python skills are essential.

Qualifications

• Bachelor’s/Master’s/PhD in Electrical Engineering or Computer Science with strong architecture background.
• Exposure to low‑level DL/LLM math and performance profiling.
• Experience 3+ years in CPU/GPU performance, kernel optimization, or HPC.
• Proficiency with C++ and Python.

Responsibilities

• Bring up and optimize performance on new Cerebras WSE generations.
• Build kernel- and end-to-end performance models for ML workloads.
• Debug kernel micro code and compiler algorithms to boost inference speed.
• Analyze runtime performance on the system and cluster.
• Develop tools to visualize performance data from the Wafer Scale Engine.

Skills

Electrical Eng

Computer Science

Deep Learning Math

Problem Solving

3+ yrs Experience

C++

Python

Education

Bachelor's/Master's/PhD in Electrical Engineering or Computer Science

Job description

Cerebras Systems, Inc. in Sunnyvale, CA, seeks a performance engineer to optimize AI workloads on the Wafer Scale Engine. You will work across hardware and software to identify bottlenecks and improve efficiency.

Responsibilities include bringing up new WSE generations, building performance models, and debugging kernel code to elevate inference speed and throughput. Strong architecture background and C++/Python skills are essential.