This list contains only the countries for which job offers have been published in the selected language (e.g., in the French version, only job offers written in French are displayed, and in the English version, only those in English).
Senior Product Architect – AI Data Center & SONiC Networking. This role has been designed as ‘’Onsite’ with an expectation that you will primarily work from an HPE office.
Job Responsibility:
Architect ultra-low-latency, lossless Ethernet fabrics supporting tens of thousands of GPUs for AI training and inference
Own the end-to-end SONiC platform architecture and fabric strategy, spanning control plane, management plane, data-plane integration, and operations at scale
Define multi-generation fabric and platform strategy across switch ASICs, NICs, SerDes capabilities, cabling, and system constraints, aligned to power, performance, and deployment realities
Own link-level and physical-layer requirements as they impact SONiC performance, including high-speed PAM4 signaling (112G/224G), error handling, and hardware/software interaction
Align SONiC architectures with next-generation GPU, NIC, and switch platforms, ensuring optimal performance across hardware and software boundaries
Define SONiC capabilities for AI and HPC workloads, including: Lossless Ethernet and RoCE
Congestion management, QoS, and ECN
Dynamic and flow-based load balancing
Drive scale, performance, and resiliency targets for SONiC-based fabrics, including fast convergence, hitless upgrades, and failure recovery
Define and enforce system-level validation criteria, including scale testing, fault injection, performance benchmarking, and upgrade scenarios
Drive telemetry, observability, and proactive reliability across SONiC and underlying hardware to enable early fault detection and rapid root-cause analysis
Lead cross-layer performance analysis and root-cause investigations spanning SONiC software, ASIC pipelines, SerDes behavior, power, and thermal interactions
Partner with ASIC vendors, system vendors, and internal platform teams to influence roadmaps and ensure SONiC readiness and scalability
Serve as a technical authority and cross-functional leader, translating architectural intent into clear product requirements and multi-generation roadmaps
Requirements:
10 plus years of experience in data center networking, AI infrastructure, or high-performance systems
Deep expertise in: SONiC architecture and internals
Large-scale Ethernet fabrics
High-speed SerDes (112G/224G PAM4) and their impact on system performance
Strong understanding of ASIC pipelines, buffering, ECMP behavior, and congestion mechanisms
Proven ability to diagnose cross-layer performance and reliability issues involving software, hardware, and physical-layer interactions
Hands-on experience with RDMA/RoCE, congestion control, and lossless Ethernet at scale
Experience with automation and tooling (Python, Ansible, Terraform) in large-scale environments
Industry certifications (e.g., CCIE, JNCIE, NVIDIA) or equivalent practical experience preferred