Principal GPU Validation Engineer
About The Position
As a Principal GPU Validation Engineer, you will architect, validate, and debug large‑scale AMD GPU systems spanning device, node, chassis, and rack-level deployments. You will define system-level test strategies for multi-GPU, multi-node accelerator platforms, ensuring correctness, performance, scalability, and reliability across hardware and software boundaries. This role is deeply technical and hands-on, involving GPU bring-up, firmware/driver interaction, networking validation (RDMA), and large-scale cluster enablement. You will directly influence product readiness and future AMD GPU platform designs by providing system-level feedback into architecture, silicon features, and validation infrastructure. You are a system thinker with deep technical instincts, capable of root-causing failures that span GPU silicon, PCIe/Infinity Fabric, networking, drivers, firmware, and orchestration layers. You are comfortable debugging issues that only emerge at scale—during long‑running workloads, high-throughput fabric stress, or multi-node synchronization scenarios. You bring: Proven technical leadership in complex GPU/accelerator environments The ability to translate low-level failures (timeouts, hangs, data corruption) into actionable root causes Strong collaboration skills across Architecture, Design, Firmware, Software, and Validation teams A track record of building robust, repeatable test infrastructure, not one-off debug scripts
Requirements
- Extensive experience in system-level development and validation for GPU or accelerator-based platforms in data center or HPC environments
- Strong understanding of: GPU architectures and software stacks (drivers, runtime, firmware interaction) GPU memory hierarchies, data movement, and synchronization Scale-up and scale-out fabrics and their performance/debug implications
- Hands-on experience with application scaling on GPU or accelerator clusters, including: Debugging scaling inefficiencies Identifying bottlenecks across compute, memory, and network fabrics
- Deep knowledge of networking and RDMA-enabled accelerator solutions, including: Fabric bring-up and validation Latency, bandwidth, and congestion analysis Debugging data corruption, packet loss, or timeout scenarios
- Strong software development skills in: C++ for performance-critical tools and validation components Python for automation, orchestration, and data analysis UNIX/Linux shell scripting for system control and diagnostics
- Proven experience building and maintaining CI/CD pipelines for: Large-scale system validation Automated regression testing Continuous integration across hardware and software changes
- Demonstrated ability to operate in a fast-paced, high-demand environment where rapid triage, decision-making, and execution are required
- Deep experience in system-level engineering with accelerators, GPUs, or complex compute platforms
Responsibilities
- Define and execute solutions for system integration and validation for AMD GPU platforms across: Single-GPU and multi-GPU nodes Scale-up fabrics (e.g., multi-GPU interconnects) Scale-out clusters and rack-level deployments
- Develop and execute solutions which drive scale up and scale out fabric, including: GPU topology validation PCIe, memory integration with the fabric stress testing Fabric related Driver, firmware, and OS interaction validation Performance and stability testing under real workloads as well Synthesizing down real world test into smaller repeatable test
- Lead deep system-level debug efforts, including: GPU hangs, resets, and error recovery scenarios at scale
- RDMA and networking-related data integrity or performance issues Multi-node synchronization and scaling failures Long-haul stability issues seen in burn-in or customer environments
- Collaborate closely with Architecture, Design, and Software Engineering teams to: Improve test hooks, telemetry, and debug visibility Provide feedback for future silicon, firmware, and platform features Ensure system-level requirements are addressed early in the design cycle
- Design and implement validation infrastructure which Leverages AI automation, including: Automation frameworks and CI/CD pipelines System health monitoring, logging, and failure triage tools AI Automation Test Pipeline
Benefits
- AMD benefits at a glance.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Principal
Number of Employees
5,001-10,000 employees
