Infrastructure Architect (GPU)

About the Client

Our Client is a publicly traded artificial intelligence (AI) and high-performance computing (HPC) infrastructure provider.

About the Role

• Lead the end-to-end hardware architecture design and planning for AI Computing Centres (AI Data Centres), encompassing compute, networking, storage, rack, power, and cooling infrastructure.
• Design and optimize GPU cluster architectures, including multi-GPU interconnects (NVLink/NVSwitch), high-speed inter-node networking (InfiniBand/RoCE), and non-blocking network topologies such as Spine-Leaf and Fat-Tree.
• Define hardware standards, technical specifications, and deployment architectures for large-scale AI and high-performance computing environments.
• Evaluate, select, and validate servers and key hardware components, including CPUs, GPUs, memory, storage, network interface cards (NICs), and power systems.
• Develop and maintain Bill of Materials (BOM), sizing models, capacity plans, and hardware configuration standards.
• Perform hardware benchmarking, performance testing, compatibility validation, and capacity planning to ensure optimal infrastructure performance.
• Provide technical leadership for hardware deployment activities, including rack installation, cabling, commissioning, power-on testing, acceptance testing, and troubleshooting.
• Collaborate with data centre, network, facilities, and operations teams to ensure successful implementation of AI infrastructure solutions.
• Manage hardware vendors and technology partners to ensure solution quality, delivery timelines, and technical compliance.
• Drive architecture reviews and recommend technology improvements to enhance performance, scalability, reliability, and cost efficiency.
• Track industry trends and emerging technologies including next-generation GPUs/NPUs, Co-Packaged Optics (CPO), liquid cooling solutions, and AI computing infrastructure innovations.
• Support the development of technical proposals, solution documentation, implementation plans, and customer presentations.

What We're Looking For

• Bachelor's Degree in Computer Science, Electronics Engineering, Communications Engineering, Automation, or a related discipline.
• Minimum 5 years of experience in server, data centre infrastructure, hardware architecture, or AI computing solutions.
• Strong understanding of server hardware architecture, including CPU, GPU, memory, PCIe, NVLink, and accelerator technologies.
• Hands-on experience with GPU-based platforms such as NVIDIA HGX and multi-GPU server architectures.
• Strong knowledge of data centre networking technologies, including InfiniBand, RoCE, Ethernet fabrics, and high-performance network design.
• Experience designing scalable, non-blocking network architectures for AI and HPC environments.
• Understanding of data centre facilities, including power distribution, cooling systems, liquid cooling technologies, rack design, and high-density deployment considerations.
• Proven experience in hardware sizing, component selection, performance testing, and infrastructure capacity planning.
• Ability to design and deliver end-to-end hardware architecture solutions and BOM recommendations.
• Strong analytical, problem-solving, vendor management, stakeholder management, and technical documentation skills.
• Experience working across cross-functional teams, including engineering, operations, facilities, and external vendors.

What's Next

If you are ready to take the next step in your career journey within a supportive global team environment-where your skills are valued and developed-this opportunity awaits your application. Click APPLY or email your resume to [email protected]

Do note that we will only be in touch if your application is shortlisted.

Robert Walters (Singapore) Pte Ltd

ROC No.: 199706961E | EA Licence No.: 03C5451

EA Registration No.: R1324990 Neha Singh