Latest AI JobsBrowse Jobs
pragmatike logo

AI Infrastructure Hardware Engineer

pragmatikeAPAC / China / Hong KongWork with our Clients
Apply on company site
Salary
Not listed
Job type
Full-time
Experience
Mid
Posted
10 Sept 2026
Status
Open · checked 7d ago

About this role

Based in Hong Kong, this hybrid AI Infrastructure Hardware Engineer role suits an autonomous server professional with fluency in both English and Mandarin. The position focuses on quality control, testing, and validation of custom HPC and GPU servers for an expanding sovereign cloud infrastructure.

Role Overview

Pragmatike is hiring an AI Infrastructure Hardware Engineer to work on-site in Hong Kong across manufacturing sites, test labs, and factory floors. The professional joining this team will take total ownership of hardware quality assurance for a European deep-tech company that designs and builds energy-efficient, sovereign cloud hardware from the ground up.

Key Responsibilities

Manage quality control on the factory floor across manufacturing locations.
Operate the hardware test lab for system bring-up, validation, and stress testing of new server hardware.
Supervise assembly and testing procedures for GPU servers prior to deployment.
Establish test procedures, acceptance criteria, and quality control checkpoints for full systems, boards, and chassis.
Troubleshoot issues relating to hardware, firmware, drivers, thermal performance, power, and PCIe buses, conducting root-cause analysis alongside design and manufacturing teams.
Run Linux-based validations for GPU compute and interconnect performance utilizing CUDA and NCCL benchmarking tools.
Log test yields, failures, and results to inform future production and hardware design choices.
Communicate daily with local suppliers and manufacturing partners using fluent English and Mandarin.

Requirements

Demonstrated practical background in testing, burn-in, bring-up, and validating HPC or GPU servers rather than only operating them.
Advanced system- and hardware-level Linux abilities.
Competence in firmware management including BIOS, UEFI, BMC, IPMI, GPU, and NIC firmware flashing and debugging.
Practical familiarity with the NVIDIA GPU software ecosystem, covering drivers, NCCL, CUDA, and multi-GPU benchmarking.
Comprehensive knowledge of server hardware components such as power systems, cooling, motherboards, networking, storage, and PCIe topology.
Background in quality control or manufacturing test environments for hardware production.
Mandatory fluency in both spoken and written English and Mandarin.
Readiness to work entirely on-site in Hong Kong within lab and factory settings.

Nice to Have

Background working with micro-datacenter or containerized deployments dealing with specific environmental, power, and thermal limits.
Knowledge of high-speed interconnect validation including NVLink, RoCE, and InfiniBand.
Programming capabilities in Python or Bash for automating test workflows and reporting.
Prior interaction with contract manufacturers or hardware supply chains within Southern China and Hong Kong.
Conversational or fluent Cantonese.

Skills and Tools

Linux
CUDA
NCCL
BIOS / UEFI
BMC / IPMI
Python
Bash
NVIDIA GPU Stack
InfiniBand
RoCE
NVLink

About the Company

Pragmatike acts as a recruitment partner for a European deep-tech enterprise creating energy-efficient, sovereign cloud infrastructure. The organization independently designs and manufactures its own containerized units, boards, and chassis as part of a vertically integrated alternative to legacy cloud services.

How to Apply

Interested candidates should apply directly on the official company careers page.

Skills and tech stack

LinuxCUDANCCLBIOSUEFIBMCIPMIPythonBashInfiniBandRoCENVLink

Questions candidates ask

Where is this job located?

The role is based in Hong Kong as a hybrid position operating directly on factory floors, manufacturing sites, and test laboratories.

What language proficiencies are required?

Both fluent English and fluent Mandarin are mandatory, non-negotiable requirements for the position.

Is this a remote or desk-based position?

No, this is explicitly not a remote or desk-based role, requiring on-site presence in Hong Kong.

What kind of hardware will the engineer be working with?

The engineer will work with high-performance computing (HPC) and GPU servers, chassis, boards, and containerized units.

What software and technology stack experience is needed?

Candidates need a strong Linux background, familiarity with firmware like BIOS/BMC, and experience with the NVIDIA GPU stack including CUDA and NCCL.

Ready to apply?

The application is on the official pragmatike careers page.

Apply on company site