ML Inference Performance Architect

Cerebras; United States; On-site; USD 100,000 - 130,000; Full time

Job summary

Cerebras is seeking engineers for the inference performance team, focusing on optimizing model inference speed and throughput. Responsibilities include building performance models, debugging kernel micro code, and developing tools for visualizing performance data.

The ideal candidate holds a degree in Electrical Engineering or Computer Science and has 3+ years of experience in CPU/GPU performance and kernel optimization. A diverse, equal opportunity environment is promoted at Cerebras, fostering inclusive teams.

Qualifications

• 3+ years of experience in Computer Architecture, CPU/GPU Performance, or Kernel Optimization.
• Strong analytical and problem-solving mindset.
• Understanding of low-level deep learning / LLM math.

Responsibilities

• Build performance models to estimate performance of ML models.
• Optimize and debug kernel micro code and compiler algorithms.
• Develop tools to visualize performance data.

Skills

Computer architecture

C++

Python

Kernel optimization

Performance profiling

Education

Bachelors / Masters / PhD in Electrical Engineering or Computer Science

Tools

CPU/GPU simulators

Job description

Cerebras is seeking engineers for the inference performance team, focusing on optimizing model inference speed and throughput. Responsibilities include building performance models, debugging kernel micro code, and developing tools for visualizing performance data.

The ideal candidate holds a degree in Electrical Engineering or Computer Science and has 3+ years of experience in CPU/GPU performance and kernel optimization. A diverse, equal opportunity environment is promoted at Cerebras, fostering inclusive teams.