below marketmost similar roles pay hereabove market
This role pays less than
65%
of similar roles. Most pay
$162,000–$235,750
— the blue band above.
At the midpoint, this role pays about
$177k
versus about
$199k
for comparable roles.
Based on 240 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
634 open roles
on FindRole.
Listed pay typically runs
$119,800–$234,700
across 612 roles with salary data.
The Senior AI Performance Architect joins the Systems Planning and Architecture team within the Azure Hardware Systems and Infrastructure organization. In this role, you will analyze frontier training and inference workloads to identify performance bottlenecks and develop analytical models, simulators, and data-analysis tools. You will evaluate trade-offs in scalability, utilization, and cost across accelerator, memory, networking, and software configurations. Key responsibilities include running targeted microbenchmarks on silicon to measure compute and communication performance, and guiding hardware-software co-design decisions for cloud-scale AI systems. Required skills include experience in computer architecture, performance modeling, and systems analysis. Preferred technical proficiencies include Python, C++, CUDA, Triton, or ROCm/HIP, alongside knowledge of transformer-based models, HBM hierarchies, and distributed execution strategies like tensor or pipeline parallelism.
Analyze frontier training and inference workloads to identify performance requirements and bottlenecks.
Develop and improve analytical models, simulators, and data-analysis tools for AI system performance.
Evaluate trade-offs in performance, scalability, utilization, capacity, and cost across hardware and software configurations.
Translate workload insights into architecture requirements by studying hardware-software interactions in large-scale systems.
Guide hardware-software co-design decisions for cloud-scale AI systems with MSI and partner teams.
Run targeted microbenchmarks on silicon to measure compute, memory, communication, and synchronization performance.
Compare analytical model predictions with silicon measurements to identify and close performance gaps.
Communicate findings and recommendations through data visualizations, technical reports, and architecture decision materials.
What we're looking for
Master's Degree in a related field and 3+ years of technical engineering experience, or Bachelor's Degree and 5+ years, or equivalent experience.
Ability to pass the Microsoft Cloud background check.
Doctorate in a related field and 3+ years of experience, or Master's and 6+ years, or Bachelor's and 8+ years (preferred).
Experience in AI systems performance, computer architecture, performance modeling, high-performance computing, distributed systems, or accelerator software (preferred).
Strong understanding of computer architecture and performance interactions among compute, memory, communication, and software (preferred).
Experience with analytical performance modeling, simulation, profiling, workload characterization, or system performance analysis (preferred).
Hands-on programming experience in Python and at least one systems or accelerator programming environment such as C++, CUDA, Triton, or ROCm/HIP (preferred).
Experience with transformer-based models, large language model training/inference, or distributed execution strategies (preferred).
CUDA
C++
Python
GPU Architecture
Deep Learning
Neural Networks
Performance Modeling
MLPerf Training
Computer Architecture
System Simulators
Performance Analysis
Drivers
DL Frameworks
CUDA
C++
C
OpenCL
MPI
NVSHMEM
OpenSHMEM
Python
IPC
Linear Algebra
Numerical Methods
Software Design
Algorithms
GPU Programming
Multi-node Communication