AI Instinct System Management Architect

Advanced Micro Devices, Inc•Santa Clara, CA

1d•Hybrid

About The Position

The AI Instinct System Management Architect will define and drive the architecture for system management and observability across AMD’s AI datacenter platforms. This role spans firmware, operating systems, rack controllers, and orchestration layers to deliver highly manageable, composable, and secure infrastructure at rack and pod scale. It combines deep technical expertise with customer engagement, owning reference designs and blueprints for deploying, integrating, and operating system management at scale.

Requirements

Expert background in systems or platform software architecture with focus on system management and server manageability.
Deep expertise in BMC firmware stacks, telemetry, inventory, alerting, and management protocols.
Strong knowledge of DMTF standards (MCTP, PLDM, SPDM, Redfish), platform security, and management networking.
Experience with PCIe, CXL, NVMe interconnects and cluster schedulers (Kubernetes, Slurm).
Proven ability to combine technical leadership with customer engagement for scalable AI datacenter deployments.

Nice To Haves

Experience with GPU/accelerator platforms
Familiarity with telemetry stacks (Prometheus, Loki, ELK, OpenTelemetry).
Knowledge of datacenter infrastructure components (racks, PDUs, power/thermal systems, fabric networking).
Contributions to open standards or open-source projects related to manageability or observability.
Strong programming skills in C/C++ and Python/Go; solid Linux systems experience.

Responsibilities

Develop a unified architecture for rack-scale and pod-scale management, ensuring seamless integration from device-level firmware to orchestration layers.
Benchmark against competitors, track emerging standards, and propose innovative features to position AMD as a leader in AI datacenter manageability.
Ensure compatibility with industry standards (Redfish, DMTF profiles) and customer monitoring stacks for observability and analytics.
Define and deliver manageability solutions (BMC/BSP, rack/pod controllers, APIs) ensuring coherent architecture from device to orchestration.
Develop standards-based interfaces and telemetry frameworks (Redfish, DMTF) for compute, storage, networking, and accelerators at scale.
Enable discovery, provisioning, firmware upgrades, and decommissioning workflows across racks and pods.
Collaborate with customers and partners to align with DCIM/ITSM environments, validate designs, and lead proof-of-concepts.
Produce architectural collateral and documentation, including reference designs, integration guides, and telemetry baselines for customer readiness.
Influence product strategy with customer-backed roadmaps optimized for AI workloads.