AI Infrastructure Hardware Engineer
At a glance
Mid-level ML Platform & Ops role at Pragmatike. Dongguan · full-time.
Pay not stated ML Platform & Ops salary
aimlroles summary, based on the employer's posting.
What you'll do
- Set quality checks across manufacturing sites and run server board bring-up, validation, and stress tests in the lab
- Check GPU servers through assembly and full-system testing before they enter deployment
- Set acceptance rules and inspection points for chassis, boards, and complete systems
- Track hardware, firmware, and driver faults through root-cause work with design and manufacturing teams
- Benchmark GPU compute and interconnects on Linux, then record test outcomes for production and design teams
What you bring
- Hands-on GPU or HPC server testing, including bring-up and validation, rather than experience operating servers alone
- Strong Linux skills and practical firmware troubleshooting across BIOS or UEFI, BMC or IPMI, and GPU or NIC firmware
- Working knowledge of NVIDIA drivers, CUDA, NCCL, and benchmarking multi-GPU systems
- Fluency in both English and Mandarin, plus experience with quality checks or testing in hardware production
Who this fits
This role suits a hands-on server engineer who is comfortable switching between a factory floor and a test lab, with ownership of quality sign-off. You must work on-site in Hong Kong and communicate with manufacturing partners and suppliers in English and Mandarin.
From the employer
Job Description
Location: Hong Kong — Hybrid (factory floor, test lab and manufacturing sites; this is not a remote or desk-based role)
Start date: ASAP
Languages: Fluent English and fluent Mandarin are both mandatory
Industry: HPC / GPU Server Manufacturing / Sovereign Cloud Infrastructure
Pragmatike is recruiting on behalf of a European deep-tech company building a sovereign, energy-efficient alternative to traditional cloud providers. Unlike most cloud players, our client designs and manufactures its own HPC infrastructure end to end — chassis, boards and containerized units — and deploys it as the hardware layer of a vertically integrated AI cloud. The company is growing fast and is expanding its manufacturing and testing capability in Hong Kong.
Professional Background:
We're looking for a Server Engineer with hands-on GPU/HPC hardware experience who has tested and validated servers, not just operated them. You should be comfortable moving between the factory floor and the test lab, with a strong Linux background and real familiarity with firmware and GPU software stacks (CUDA, NCCL). You take ownership of quality: if a server ships, it is because you signed off on it.
Startup or hyper-growth experience is a strong plus. Autonomy, rigor and problem-solving are essential.
Your Responsibilities:
Own quality control across the client's manufacturing sites, on the factory floor.
Run the hardware test lab: board and system bring-up, validation and stress testing of new server builds.
Oversee the assembly and end-to-end testing of GPU servers before they are deployed into the client's sovereign cloud.
Define and maintain test procedures, acceptance criteria and QC checkpoints for chassis, boards and full systems.
Diagnose hardware, firmware and driver-level issues (BIOS/BMC, GPU firmware, PCIe, thermals, power) and drive root-cause analysis with the design and manufacturing teams.
Validate GPU compute and interconnect performance under Linux using CUDA, NCCL and related benchmarking tools.
Document failures, yields and test results, and feed findings back into hardware design and production processes.
Coordinate day to day with manufacturing partners and local suppliers in both English and Mandarin.
What You Bring:
Proven hands-on experience testing and validating GPU/HPC servers (bring-up, burn-in, validation), not only using them.
Strong Linux skills at the system and hardware level.
Firmware knowledge: BIOS/UEFI, BMC/IPMI, GPU and NIC firmware, flashing and troubleshooting.
Working knowledge of the NVIDIA GPU stack: drivers, CUDA, NCCL, and how to benchmark and diagnose multi-GPU systems.
Solid understanding of server hardware: motherboards, PCIe topology, power, cooling, storage and networking components.
Experience with QC or manufacturing test processes in a hardware production environment.
Fluent English and fluent Mandarin (both non-negotiable).
Willingness to work on-site in Hong Kong, on the factory floor and in the lab.
Nice To Have:
Experience with containerized or micro-datacenter deployments (power, thermal and environmental constraints).
Familiarity with high-speed interconnects (InfiniBand, RoCE, NVLink) and their validation.
Scripting skills (Bash/Python) to automate test sequences and reporting.
Exposure to hardware supply chains or contract manufacturers in Hong Kong / Southern China.
Cantonese.
Why Join Us:
Our client is one of the very few cloud companies that builds its own hardware from the board up. Their mission is to deliver a sovereign, energy-efficient, high-performance alternative to traditional cloud providers, and the servers you test and validate are the physical foundation of that platform.
Expect real ownership from day one, a fast-paced environment that values autonomy and problem-solving, and the chance to build something that doesn't exist yet.
Pragmatike• is dedicated to a fair, transparent, and inclusive recruitment process. We ensure that no applicant is discriminated against based on age, disability, gender, gender identity or expression, marital or civil partner status, pregnancy or maternity, race, religion or belief, sex, or sexual orientation. In accordance with the General Data Protection Regulation (GDPR), your personal data will be processed lawfully, fairly, and securely. We collect and use your personal data solely for recruitment purposes, including sharing it with our client(s) for employment consideration. You have the right to request access, correction, or deletion of your data at any time. We are committed to maintaining the confidentiality and security of your information throughout the recruitment process.