TheDomain Architect - AI Networkacts as the primary technical authority for the physical and logical lifecycle ofhigh-performance interconnect fabricsacross diverse client environments, who bridges the gap between architectural design and hands-on execution.
You are a "doer" who is as comfortable configuringAdaptive Routing on leaf and spine switchesin the CLI as you are explaining that configuration to a C-level client.
As a System Integrator, we do not simply manage a static cloud; we design and deliver bespoke, high-scale AI factories for the world's leading enterprises.
In this role, you will define the "Gold Standard" fornetworkinfrastructure, moving beyond single-switchmanagement to architect repeatable, scalable, and automatedlosslessfabrics.
You will serve as the technical lead for NVIDIA Cloud Provider (NCP) and private enterprise AI cloud deployments, owning the "Network" in the critical "Compute-Network-Storage" triad.
In this role, you will operate with a***** splitbetween delivering (60) complex AI infrastructure and providing Pre-Sales (40) Subject Matter Expertise (SME).
You will lead the physical provisioning ofInfiniBand and Spectrum-X Ethernet fabrics for NVIDIA SuperPOD, NVIDIA BasePOD, and Cisco AI Factory environments, ensuring our clients receive "Day 2" ready AI factories, while assisting the sales team in defining the scope and cost of future deployments.
Key Responsibilities
1. Delivery & Implementation (60%)
High-Performance Fabric Design:
Architect and deployNon-Blocking Fat-Tree (Clos)topologies usingNVIDIA Quantum-2 (NDR) / Quantum-3 (XDR) InfiniBand switchesandSpectrum-4 Ethernet switches.
ImplementRail-Optimisednetwork designs to ensure GPU-to-GPU traffic is perfectly aligned with compute node PCIe trees, minimising latency for NCCL collectives.
ConfigureAdaptive Routing,Congestion Control, andQuality of Service (QoS)to prevent "Head-of-Line Blocking" and ensure lossless data delivery.
In-Network Computing & Offload Strategy:




InfiniBand (Switch-Based):Enable and tuneSHARP (Scalable Hierarchical Aggregation and Reduction Protocol)on Quantum switches to offload collective operations (AllReduce, ReduceScatter) from the GPU to the switch silicon.
Ethernet (Endpoint-Based):Architect high-performance Ethernet AI fabrics utilisingSmartNIC/DPU offloading(e.g.,NVIDIA BlueField-3) to accelerate collective operations and isolate management traffic from the data path.
Future Stacks:Evaluate and roadmap emergingUltra Ethernet Consortium (UEC)standards and hardware capabilities to transition In-Network Collectives from proprietary InfiniBand to open Ethernet standards.
TuneRoCEv2and adaptive routing algorithms on Ethernet fabrics and SmartNICs to approximate InfiniBand-like lossless behaviour.
Fabric Management & Telemetry:
DeployNVIDIA UFM (Unified Fabric Manager)to manage InfiniBand subnets andNetQfor Ethernet telemetry.
Monitor fabric health to identify "Slow Receivers" and link degradation in real-time.
ImplementPTP (Precision Time Protocol)for nanosecond-level clock synchronisation across the entire cluster.
Layer 1 Precision:
Oversee the physical cabling strategy, defining strict cable schedules (DAC vs. AOC vs. OSFP transceivers) to meet signal integrity requirements over specific distances.
Validate the "Link Budget" to ensure optical loss falls within acceptable limits for 400G/800G links.
Technical Scoping & Sizing:
Calculate the requiredBisectional Bandwidthfor client workloads (e.g., Training requires 1:1 non-blocking, Inference may tolerate 3:1 oversubscription).
Produce accurateLabour Estimates (LOE)for complex network integrations.
BoM Validation:
Own the Network Bill of Materials (BoM), validating that every transceiver, cable, and switch is on theNVIDIA/OEM HCL.
Manage the complexities of breakout cables (e.g., 800G to *****G) and connector types (OSFP, QSFP112)



to prevent on-site installation failures.
Educate clients on the difference between standard Enterprise Ethernet (Lossy, TCP-based) and AI Fabric (Lossless, RDMA-based).
Design integration points between the high-speed "Backend" AI fabric and the client's existing "Frontend" management network (BGP/EVPN handoffs).
Deep expertise inNVIDIA Quantum InfiniBandswitch architecture and management viaMLNX-OSand UFM.
Mastery ofSpectrum-XEthernet platforms runningNVIDIA Cumulus LinuxorSONiC (Software for Open Networking in the Cloud).
Proficiency in configuringRoCEv2(RDMA over Converged Ethernet), includingPFC(Priority Flow Control) andECN(Explicit Congestion Notification).
Architectural understandingofNVIDIA BlueField DPUandDOCAuse cases for offloading network functions.
Network Automation (NetDevOps):
Proficiency inPythonandAnsiblefor switch configuration management.
Experience withInfrastructure as Code (IaC)principles to treat network config as software.
Fabric Telemetry:
Hands-on experience withUFM,Prometheus, andGrafanafor visualising fabric congestion and "tail latency.
"
Multi-Vendor Exposure:Experience withArista (EOS)for high-performance AI Ethernet orCisco (Nexus Dashboard)for AI infrastructure.
Routing Protocols:Solid understanding ofBGP,EVPN, andVXLANfor multi-tenant isolation.
Compute Integration:Understanding of how the network interacts with the host OS (IPoIB, Netlink) and optimisation ofGPU Direct RDMA (GDR)andNCCL communication patterns.
Scale-Across: Extending an InfiniBand fabric across multiple data centrehalls (NVIDIA Metro-X).
Certifications:
NVIDIA-Certified Skilled: AI Networking (NCP-AIN)
NVIDIA-Certified Associate: AI Infrastructure and Operations (NCA-AIIO)
Fabric Performance:Achieving >95% effective bandwidth efficiency onNCCL-testbenchmarks across the full cluster.
Deployment Accuracy:Zero "Layer 1" rework required due to incorrect transceiver/cable specification in the BoM.
Stability:Zerotopology or design-induced link instability(e.g., cabling mismatches) in production.
#J-*****-Ljbffr

📌 Domain Architect - Ai Network (New South Wales)
🏢 World Wide Technology
📍 New South Wales