| Microsoft Careers

Microsoft

Quick summary

Work type: On-site
Location: —
Posted: 46 days ago

Market check

Salary context

How this pay compares to similar roles

Similar $180k

$111k most similar roles pay here $248k

This listing doesn't post a salary. Most similar roles pay $152,150–$208,800.

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 728 open roles on FindRole.

Listed pay typically runs $119,800–$234,700 across 664 roles with salary data.

Most-posted roles

View all roles at Microsoft

At a glance

TL;DR · | Microsoft Careers

Apply Now Log in to save

As a Research Intern within Microsoft’s Azure Hardware and Systems Infrastructure team, you will engage in cutting-edge hardware/software co-design for large-scale AI models, focusing on optimizing CPU, GPU, and networking infrastructure for the Maia accelerator ecosystem. Your day-to-day responsibilities include identifying system stress points, proposing architectural innovations, and developing methodologies to evaluate their impact through workload characterization, modeling, and benchmarking. This role requires proficiency in frameworks like PyTorch, SGLang, Dynamo, CUDA, and Triton, alongside a deep understanding of GPU and AI architectures, memory hierarchies, and CPU/server configurations. You must also possess hands-on experience with performance analysis tools, trace generation, event monitoring, and roofline modeling to address complex system architecture challenges in a fast-paced environment.

Skills

PyTorch CUDA Dynamo SGLang Triton GPU CPU PCIe roofline_performance_modeling performance_analysis profiling benchmarking trace_generation event_monitoring instrumentation Kubernetes Azure Python PostgreSQL

What you'll do

Identify system stress points in CPU, GPU, and networking infrastructure for AI workloads.
Propose novel architectural ideas to optimize central processing units (CPUs) and GPUs.
Create methodologies using workload characterization, modeling, and benchmarking techniques.
Evaluate the effectiveness of proposed architectural changes through detailed simulations.
Apply performance analysis tools and methodologies to identify and resolve bottlenecks.

What we're looking for

Currently enrolled in a PhD program in Computer Science or related STEM field.
At least 1 year of experience with performance analysis tools, optimization, and modeling.
Proficiency with AI frameworks like PyTorch and programming models such as CUDA and Triton.
Deep understanding of GPU and AI architectures including memory hierarchies and interconnect properties.
Familiarity with CPU/server architectures, PCIe topologies, and accelerator/NIC/peripheral demand.
Hands-on experience in benchmarking, profiling, identifying performance bottlenecks, and optimization.

| Microsoft Careers

Microsoft

US 46 days ago

Azure Python C# JavaScript R Terraform Bicep Azure Functions Docker API Management Azure Cognitive Services Azure OpenAI Azure AI Search Vector Indexes Azure Document Processing Infrastructure as Code CI/CD

Save