Principal Software Engineer, Performance

Microsoft

Quick summary

Work type: On-site
Location: Mountain View, CA
Salary: $142,800–$274,800 / yr
Posted: 18 days ago
Closes: Dec 6, 2026
Nearby: 99+ roles within 25 mi

Market check

Salary context

Competitive pay

How this pay compares to similar roles

Similar $198k

This role $209k

$127k most similar roles pay here $291k

This role pays more than 64% of similar roles. Most pay $174,200–$221,775 — the shaded band above. At the midpoint, this role pays about $209k versus about $198k for comparable roles.

Based on 239 similar postings.

Employer

About Microsoft

Microsoft Corporation is a global technology leader producing software, hardware, and cloud services including Windows, Office 365, Azure cloud platform, Xbox gaming, and Surface devices. Industry: Software & Cloud Computing

Microsoft currently has 622 open roles on FindRole.

Listed pay typically runs $119,800–$234,700 across 571 roles with salary data.

Most-posted roles

View all roles at Microsoft

At a glance

TL;DR · Principal Software Engineer, Performance

Role Posting Log in to save

As a Principal Software Engineer on the Artificial Intelligence Cloud Inference team at Microsoft, you will work on optimizing the performance of large language models (LLMs) like those from OpenAI across various hardware platforms, including GPUs and custom silicon. Your day-to-day responsibilities include benchmarking LLMs for performance, debugging and optimizing code at multiple levels, developing software tools to monitor and enhance system efficiency, and collaborating with internal and external partners. You will need expertise in high-performance computing, GPU architecture, and deep learning frameworks such as PyTorch or TensorFlow, along with proficiency in CUDA and ROCm. This role involves tackling complex technical challenges to improve the end-to-end performance of AI models used in major Microsoft products like Office and Bing, contributing significantly to Azure's cost efficiency goals.

Skills

Python C++ CUDA ROCm PyTorch TensorFlow ONNX_Runtime NVIDIA_GPUs AMD_GPUs Maia_silicon Performance_Benchmarking GPU_Profiling_Tools CI/CD Azure Linux

What you'll do

Identify and drive improvements to end-to-end inference performance of LLM models.
Benchmark performance on Nvidia/AMD GPUs and Microsoft silicon for LLMs.
Optimize and monitor performance at various levels from model to system.
Build software tooling to enable insights into performance opportunities.
Design, implement, and test functions or components for AI/DNN/LLM frameworks.
Reduce complexity of key components/pipelines to improve system efficiency.

What we're looking for

4+ years of practical experience in high-performance application development and performance debugging on CPUs/GPUs.
Strong foundation in software engineering principles, computer architecture, and GPU architecture.
Experience in end-to-end performance analysis and optimization of state-of-the-art LLMs and HPC applications.
Proficiency with GPU profiling tools and familiarity with CUDA, ROCm, Triton.
Hands-on experience with DNN/LLM inference and one or more DL frameworks like PyTorch, TensorFlow, or ONNX Runtime.

Hybrid

Save