
GPU Software System Management Engineer
Job Details
Job Description
The Role and Impact
As a Cloud Software Development Engineer, you will design, develop, validate, and debug software solutions that optimize Intel hardware features for cloud environments and deployment models. Your day-to-day work will focus on enabling seamless integration across firmware, drivers, OS, middleware, SDKs, frameworks, and applications, ensuring maximum performance and value for Intel products. Through your contributions, you will play a key role in advancing cloud, hybrid-cloud, and on-premises solutions that empower customers to achieve their goals.
Business group
You will be joining the Data Center Group (DCG), a leading organization at Intel focused on developing cutting-edge solutions that power the world's digital infrastructure. The System Management Software team works at the intersection of technology and innovation, enabling Intel products to deliver exceptional performance and reliability for diverse cloud ecosystems. Together, we drive innovation in data services, AI/ML, and open-source development, supporting Intel's mission to transform the future of computing.
Key Responsibilities
- Design, develop, and optimize software solutions tailored to Intel hardware for cloud, hybrid-cloud, and on-premises deployment models.
- Collaborate with cross-functional teams to investigate and debug issues across firmware, drivers, OS, middleware, SDKs, and frameworks.
- Analyze partner software stacks and optimize full-stack software solutions for enhanced performance and usability.
- Develop and validate reference cloud platform software to support customer designs and maximize Intel product value.
- Contribute to specific technology domains, including data services, AI/ML, and open-source projects, ensuring alignment with industry trends and customer needs.
- Document and communicate findings, processes, and solutions effectively to stakeholders.
Qualifications
Mimimum Qualification
- 3+ years of software product development experience, preferably in system software, GPU/accelerator software, HPC, data-center infrastructure, or distributed systems.
- 3+ years of Linux system software development experience, with strong understanding of Linux OS architecture, processes, scheduling, memory management, virtual memory, device interfaces, IPC, and hardware/software interaction.
- Strong C/C++ programming skills with experience developing production-quality system software. Experience with Python, Bash, or Go is an added advantage.
- Experience with GPU system management and device-management APIs/frameworks, such as Intel Level Zero/Sysman, Intel XPU Manager (XPUM/XPU-SMI), NVIDIA NVML/DCGM/nvidia-smi, AMD SMI/ROCm, or equivalent technologies.
- Strong understanding of GPU/accelerator architecture and system architecture, including PCIe, memory hierarchy, multi-GPU systems, device topology, virtualization, and CPU/GPU interaction.
- Experience developing or integrating GPU manageability capabilities, including one or more of
Device discovery and inventory, Telemetry and monitoring, GPU health monitoring, RAS/error reporting, Diagnostics and system validation, Firmware management/update, GPU configuration and policy management, Crash/debug log collection, Device reset/recovery and survivability
- Experience with data-center GPU node and cluster manageability, including monitoring and managing GPUs across multi-GPU and multi-node environments.
- Understanding of cluster and cloud-native infrastructure, such as Kubernetes, GPU device/DRA drivers, container runtimes, Prometheus, Grafana, and OpenTelemetry. XPUM's architecture specifically includes Kubernetes integration and OpenTelemetry/Prometheus-based telemetry flows.
- Experience or strong understanding of MULTI-scale GPU systems, including
Scale-up and scale-out GPU topology, GPU/fabric/interconnect management,
Node-level versus cluster-level management, In-band and out-of-band management concepts, System health and telemetry aggregation, Large-scale GPU fleet monitoring and diagnostics
- Strong Linux debugging skills, including GDB, core dump/segmentation fault analysis, memory leak debugging, multithreading/concurrency debugging, performance analysis, and system-level troubleshooting.
- Experience with multithreaded and concurrent software development, with a solid understanding of synchronization, scalability, reliability, and performance.
- Experience with component, system, and end-to-end validation, preferably involving GPU hardware/software stacks and multi-GPU configurations.
Preferred Qualifications
- Experience with HPC and AI/ML infrastructure, including distributed AI/HPC workloads and accelerator-based systems.
- Understanding of MPI and distributed/parallel programming, particularly for multi-node GPU workloads and system diagnostics.
- Knowledge of GPU interconnect technologies and topology concepts such as PCIe, NVLink/NVSwitch, UALink or equivalent scale-up/scale-out interconnects.
- Experience with GPU RAS, telemetry, health monitoring, error handling, hardware diagnostics, and fault recovery.
- Experience integrating system-management software with Kubernetes, Slurm, Prometheus/Grafana, OpenTelemetry, Ansible, or similar orchestration, observability, and provisioning frameworks.
- Experience developing APIs, libraries, CLI tools, agents/daemons, exporters, or management services used by infrastructure and system administrators.
- Experience working with open-source software, including upstream contribution, code review, CI/CD, issue triage, and community collaboration.
- Experience working on large-scale software projects involving hardware, firmware, drivers, system software, validation, and infrastructure teams.
Job Type
Experienced Hire
Shift
Shift 1 (India)
Primary Location
India, Bangalore
Additional Locations
Posting Statement
All qualified applicants will receive consideration
Never pay to get work. If a listing asks for a fee, it is a scam. The ten signs →