Hardware Systems Engineer - AI Hardware

Cloudflare•San Francisco, CA

About The Position

Cloudflare’s Hardware Engineering team researches, develops, tests, and deploys equipment for its global network. The Hardware Systems Engineer will work with Engineering, Product, Performance, and Hardware Sourcing teams to design, qualify, and maintain Cloudflare’s worldwide fleet of GPU servers and AI infrastructure. The role contributes to end-to-end infrastructure decisions spanning compute, networking, storage, power, and cooling. It evaluates and qualifies AI hardware platforms and rack-scale GPU systems, performs hands-on testing and analysis, works with ODM/OEM vendors, translates AI workload requirements into infrastructure specifications, and drives solutions across hardware, firmware, software, and operations. The engineer also develops qualification and automation tooling, monitors developments in AI hardware, and communicates findings through technical writing and internal presentations. The position is available in Austin, Texas, or Lisbon, Portugal. Lisbon-based hires have an estimated annual salary of €54,000–€68,000. The role is eligible to participate in Cloudflare’s equity plan. Applicants who reach the offer stage may be asked to attend an in-person interview at a Cloudflare office or hub. The position may require authorization to access technology controlled under U.S. export laws.

Requirements

Bachelor’s, Master’s, or PhD degree in Computer Engineering, Electrical Engineering, or Computer Science.
Experience in AI infrastructure, systems engineering, or hardware platform engineering.
First-principles understanding of AI accelerator architecture, including compute pipelines, memory hierarchy, interconnect topology, and workload-to-hardware mapping.
Hands-on experience qualifying AI server platforms, including running NCCL tests, profiling collective operations, debugging NVLink errors, and troubleshooting PCIe lane configurations.
Understanding of rack-scale systems, including high-density power, power distribution, cooling, high-speed networking, or physical-layer constraints.
Systems-level familiarity with AI/ML frameworks such as vLLM or SGLang and understanding of their hardware demands and bottlenecks.
Familiarity with AI server form factors such as HGX, OAC, OAM, UBB, or open rack standards and their tradeoffs.
Scripting and automation skills in Python, Bash, or Go, and Linux system administration skills.
Strong written communication skills, including the ability to write technical specifications and vendor requirements documents.
Knowledge of performance benchmarking and experience testing, measuring, or simulating hardware performance.
Knowledge of Redfish, IPMI, and server remote-management protocols, and experience with firmware updates.
Ability to use AI tools and automation to accelerate hardware qualification while improving efficiency and quality.
Organizational skills and interest in managing cross-functional technical projects.
Self-motivation, initiative, and ability to move work forward in ambiguous environments without close supervision.
Ability to make and clearly explain defensible technical decisions with incomplete information.

Nice To Haves

Experience with major AI server ODMs/OEMs and their qualification and customization processes.
Exposure to hyperscaler or cloud infrastructure design, including GPU capacity provisioning, allocation, and operations at scale.
Experience with RDMA networking, RoCEv2 congestion control, or InfiniBand fabric management.
Understanding of AI workload patterns, including LLM training versus inference, model/tensor/pipeline parallelism, and associated memory and bandwidth demands.
Familiarity with open-compute hardware programs or OCP specifications.
Experience optimizing performance through hardware/software codesign.
Experience with kernel development, compiler optimization, or driver development.
Experience with observability and monitoring tools such as Prometheus and Grafana, including observing trends over time.
Experience with software development tools and processes such as Git, TeamCity, and Jira.
Experience in a hyperscale cloud infrastructure role.
Knowledge of server hardware fault debugging and ability to work with sourcing teams and vendors to improve quality.

Responsibilities

Contribute to technical requirements and end-to-end delivery of AI infrastructure, including compute, networking, storage, power, and cooling.
Evaluate, select, and qualify AI hardware platforms such as GPUs, custom ASICs, SmartNICs, and storage based on workload requirements and first-principles analysis.
Design and validate rack-scale GPU systems, including multi-node topologies, high-speed interconnects, power distribution, and thermal management.
Conduct hands-on hardware qualification, including benchmarking, stress testing, failure-mode analysis, and vendor root-cause analysis.
Assess and qualify ODM/OEM server platforms and work directly with vendors through the technical qualification cycle.
Translate AI product and workload requirements into precise infrastructure specifications.
Identify infrastructure-stack gaps and drive solutions across hardware, firmware, software, and operations.
Track developments in AI hardware and provide informed recommendations.
Build tooling and automation for qualification workflows, performance analysis, and operational visibility.
Communicate results and updates through blog posts and internal talks.
Manage cross-functional technical projects and ensure architectural designs align with delivery schedules.

Benefits

Medical/Rx insurance
Dental insurance
Vision insurance
Flexible Spending Accounts
Commuter Spending Accounts
Fertility and family-forming benefits
On-demand mental health support and Employee Assistance Program
Global Travel Medical Insurance
Short- and long-term disability insurance
Life and accident insurance
401(k) Retirement Savings Plan
Employee Stock Participation Plan
Flexible paid time off covering vacation and sick leave
Leave programs, including parental, pregnancy health, medical, and bereavement leave
Equity plan participation