Performance Modeling Software Engineer
Roles and Responsibilities
* The candidate will be working on a performance characterization infrastructure for our Client workloads.
* In particular, the candidate will be developing/maintaining a production service codebase which includes understanding of Client models and their performance, interacting with various Client software stacks (frontend to backend), and utilizing Client infrastructure (storage, memory, compute, and networking).
* The characterization covers basic compute/memory/network analysis to estimating the end-to-end latency with an architecture-specific analytical performance model.
* Fluency in Python, experience with coordinating multiple stacks (frontend to backend) is a plus.
* Good understanding of system performance analysis: Roofline analysis, characterizing compute, memory, and networking loads of a workload.
* Good understanding of large-scale infrastructure code: interaction between storage, memory, compute, and networking services
* Familiarity with machine learning and its framework (Caffe2, PyTorch) is a plus.
* Team player with excellent communication and collaboration skills.
Discover More AI Jobs:
- Address Menlo Park, California