Platform Architect, Server Infrastructure

Description

Ready to revolutionize AI datacenter thermal control solutions and deployments? Phononic is the place for you. We are seeking a results-oriented individual to join our Infrastructure Solutions organization to deliver industry leading innovative products in support of revolutionizing cooling solutions for AI datacenters.

About the role

We are seeking a highly experienced Platform Architect to lead the design and optimization of next-generation AI infrastructure platforms, with a core focus on thermal management and cooling technologies for optical interconnects, co-packaged optics (CPO), and GPU-based servers.

Responsibilities

• Define end-to-end thermal architecture strategies for GPU servers, optical interconnects, and CPO-based systems
• Develop system-level approaches to balance performance, heat dissipation, reliability, and energy efficiency
• Design and optimize solutions for: High-power GPUs and accelerators, Dense optical I/O (pluggable and co-packaged optics), Rack- and cluster-level thermal constraints
• Optimize cooling strategies for high-density AI workloads and optical bandwidth scaling
• Analyze and improve thermal resistance, junction temperatures, and cooling efficiency
• Lead design and evaluation of advanced cooling approaches, including: Air cooling (high-performance heatsinks, airflow optimization), Liquid cooling (direct-to-chip, cold plates), Immersion cooling and emerging techniques
• Optimize optical & CPO thermal integration, addressing challenges like heat dissipation in tightly integrated photonics + silicon environments, thermal impact on optical performance and signal integrity, and reliability of photonic components under thermal stress
• Collaborate with silicon photonics teams to co-design thermal-aware optical packaging architectures
• Design GPU server platforms optimized for thermal efficiency, including multi-GPU configurations and interconnect density, Power delivery and cooling integration, Airflow and liquid loop design
• Drive innovations in rack-level and data center-level cooling, including high-density rack (>50–100kW) thermal strategies, Integration with facility cooling systems, Optimization for power usage effectiveness (PUE)
• Optimize system performance under thermal constraints, including thermal throttling behavior and workload efficiency
• Ensure thermal considerations are integrated into interconnect design decisions, including Optical vs. electrical tradeoffs, Placement of optics (pluggable vs. co-packaged)
• Drive alignment between thermal design, system architecture, and product roadmap
• Influence vendor selection across cooling systems, optics, and server platforms
• Partner with teams across: Silicon and ASIC design, Photonics engineering, Mechanical and thermal engineering, Data center infrastructure

Qualifications

• Required: Bachelor’s or Master’s degree in Mechanical Engineering, Electrical Engineering, Physics, or related field
• Required: 8–12+ years of experience in system architecture, thermal engineering, or hardware platform design
• Required: Deep expertise in: Thermal management and cooling technologies for high-performance systems, GPU servers, AI/HPC infrastructure, or data center platforms, Optical interconnects and high-speed systems (1.6T+)
• Required: Strong understanding of: Heat transfer (conduction, convection, liquid cooling systems), Power density and thermal constraints in modern compute systems, Tradeoffs across performance, cooling, cost, and reliability

Preferred Experience

• Preferred: Working with hyperscalers or large-scale AI deployments
• Preferred: Experience with co-packaged optics (CPO) and silicon photonics thermal challenges
• Preferred: Hands-on design experience with: Liquid cooling systems and cold plate design, High-density rack and cluster cooling
• Preferred: Knowledge of: Optical module thermal constraints and reliability, Data center infrastructure (HVAC, liquid loops, facility integration)

Key Skills

• Deep expertise in thermal systems and heat management
• System-level thinking across compute, optics, and cooling
• Cross-disciplinary collaboration (mechanical, electrical, photonics)
• Strong problem-solving in high-density, high-power environments