below marketmost similar roles pay hereabove market
This role pays more than
71%
of similar roles. Most pay
$169,550–$220,800
— the blue band above.
At the midpoint, this role pays about
$209k
versus about
$195k
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
694 open roles
on FindRole.
Listed pay typically runs
$119,800–$234,700
across 651 roles with salary data.
The Principal Supercomputing Operations Software Engineer joins the Artificial Intelligence and High Performance Computing organization to serve as the technical authority for interconnect fabric operations across flagship AI supercomputing environments. This role involves leading complex, high-severity fabric incidents, performing deep multi-layer systems debugging, and defining operational strategies to ensure GPU availability and training stability. The engineer will architect automation, telemetry, and diagnostics while authoring authoritative playbooks and escalation models. Key technical requirements include proficiency in coding languages such as C, C++, C#, Java, JavaScript, or Python, alongside experience with InfiniBand, Subnet Manager, PCIe, firmware, drivers, and OS layers. The work focuses on solving reliability challenges within large-scale distributed systems to support frontier AI training, scientific computing, and large-scale distributed simulations.
Serve as the technical authority for InfiniBand and GPU interconnect fabric operations across large-scale AI supercomputing environments.
Lead and orchestrate complex, high-severity fabric incidents including detection, triage, mitigation, recovery, and root cause analysis.
Perform deep multi-layer systems debugging across InfiniBand, Subnet Manager, GPU interconnect, PCIe, firmware, drivers, and OS layers.
Define reliability models and failure domains to identify recurring patterns and implement systemic prevention mechanisms.
Author authoritative technical support guides, playbooks, and escalation frameworks adopted across the organization.
Architect and drive automation, telemetry, and diagnostics to improve observability and mean time to mitigation.
Influence engineering direction and set operational standards for interconnect fabrics across the platform.
What we're looking for
Bachelor's Degree in Computer Science or related technical field and 6+ years of technical engineering experience with coding in languages like C, C++, C#, Java, JavaScript, or Python.
Ability to pass the Microsoft Cloud Background Check.
6+ years of experience operating large-scale distributed systems, high-performance computing (HPC), or artificial intelligence (AI) infrastructure in production environments (preferred).
Demonstrated ownership of mission-critical production infrastructure with direct impact on service availability, GPU workloads, and customer SLAs (preferred).
Strong Linux systems knowledge with experience debugging low-level infrastructure issues across operating systems, drivers, and services (preferred).
Proven ability to reason across hardware, firmware, drivers, and software stacks to diagnose and resolve complex production issues (preferred).
Master's Degree in Computer Science or related technical field and 8+ years of technical engineering experience with coding in languages like C, C++, C#, Java, JavaScript, or Python (preferred).
Python
C
C++
PyTorch
CUDA
High-Performance Computing
Large Language Models
Deep Learning
AI/ML
Cloud Native
System Design
Software Development Life Cycle
Telemetry
Agentic AI
GPU Programming
Kubernetes
Go
Python
Linux
AWS
EKS
Terraform
Argo CD
Helm
Prometheus
Grafana
PyTorch
Ray
vLLM
SGLang
TensorRT-LLM
CUDA
NVLink
NVSwitch
NCCL
EFA
DCGM
Containers
HPC
C++
Python
C
Linux
ARM
PowerPC
x86
TCP
UDP
Machine Learning
Unit Testing
Performance Optimization
Computer Architecture
Networking Protocols
Hardware-in-the-loop