About the job
Microsoft's Hardware Systems organization is developing AI-native silicon and system-level solutions to power the next generation of frontier AI models. The MAIA platform combines custom accelerators, advanced networking technologies, large-scale distributed infrastructure, and cloud-scale software to deliver industry-leading AI training and inference capabilities. The Platform Systems Engineering (PSE) team is seeking a Sr. AI Network Systems Engineer to develop, integrate, validate, and deploy networking solutions for next-generation MAIA AI systems.
Responsibilities
Develop and integrate scale-up and scale-out networking solutions for MAIA AI training and inference systems.\\\\nTranslate AI/ML workload requirements into system-level networking requirements.\\\\nEvaluate architecture tradeoffs across performance, reliability, power, scalability, serviceability, and cost.\\\\nDrive integration, bring-up, and qualification of switches, NICs, PHYs, MAC/PCS, SerDes, optics, cables, and backplane technologies.\\\\nCollaborate across silicon, firmware, hardware, software, and OS teams to resolve complex integration issues.\\\\nPartner with technology vendors, ODMs, and CMs through qualification and production readiness.\\\\nDevelop end-to-end validation strategies covering functionality, performance, scale, interoperability, reliability, and stress.\\\\nCharacterize AI fabric performance including bandwidth, latency, link utilization, congestion behavior, and communication efficiency.\\\\nDevelop validation methodologies for rack-scale and cluster-scale AI infrastructure.\\\\nDrive qualification of next-generation electrical and optical networking technologies, including high-speed SerDes, copper interconnects, optical modules, and high-density connectivity.\\\\nEvaluate technologies such as PAM4 SerDes, DAC/AEC, OSFP/QSFP-DD, advanced optical interconnects, LPO/LRO, CPO, and silicon photonics.\\\\nAnalyze system-level tradeoffs across performance, power, reliability, signal/link integrity, and scalability.\\\\nLead root-cause analysis of complex networking issues across PHY, SerDes, MAC/PCS, switches, NICs, firmware, and system software.\\\\nDevelop and improve telemetry, diagnostics, validation automation, and network health monitoring.\\\\nUse lab characterization and system telemetry to drive issues to resolution and improve platform reliability.
Qualifications
Minimum
Master's Degree in Electrical Engineering, Computer Engineering, Mechanical Engineering, or related field AND 3+ years technical engineering experience OR Bachelor's Degree in Electrical Engineering, Computer Engineering, Mechanical Engineering, or related field AND 5+ years technical engineering experience OR equivalent experience.5+ years of experience developing or validating networking, compute, storage, or accelerator hardware systems.5+ years of experience designing, integrating, validating, or troubleshooting Ethernet-based networking infrastructure, including switches, NICs, optics, PHYs, or high-speed SerDes technologies.5+ years of experience supporting AI, HPC, cloud, or large-scale data center infrastructure deployments.
Preferred
Experience with GPU, FPGA, TPU, or custom AI accelerator systems supporting AI/ML workloads. Experience with scale-up and scale-out AI networking and distributed AI infrastructure. Electrical networking expertise in areas such as 112G/224G SerDes, PAM4, MAC/PCS, FEC, link training, signal integrity, DACs, and AECs, or optical expertise in areas such as OSFP/QSFP-DD, AOCs, optical transceivers, optical link characterization, and diagnostics. Familiarity with optical technologies such as DR4/DR8/FR4, optical link budgets, OMA, TDECQ, receiver sensitivity, LPO, LRO, CPO, or silicon photonics. Understanding of AI workload communication patterns, network collectives, traffic profiles, and congestion behavior. Familiarity with IEEE Ethernet, OIF, CMIS, and emerging AI networking technologies. Experience with Linux, telemetry, diagnostics, scripting, test automation, and network qualification. Experience with hyperscale datacenter infrastructure and hardware development from prototype through production.