Senior AI Infrastructure Engineer - High Performance Compute

Newbridge Alliance Pte Ltd; Singapore; On-site; SGD 180,000 - 240,000; Full time

Job summary

Confidential is building a greenfield GPU cloud to run AI inference and real-time rendering. You will architect, deploy, and operate thousands of GPUs, bare metal and orchestration, delivering as a service for customers with zero legacy.

This role focuses on high-performance CUDA workloads, driver stacks, networking (InfiniBand/RoCE) and multi-tenancy. You will build automation for provisioning, monitoring, and supporting AI teams with GPU workloads.

Qualifications

• 4+ years in GPU Systems / HPC / AI Infrastructure.
• Hands-on with NVIDIA GPUs, CUDA, driver stack and Kubernetes.
• Strong Linux and networking fundamentals; high-performance fabrics.

Responsibilities

• Architect, deploy, and operate large-scale GPU infrastructure - bare metal + orchestration
• Optimize GPU performance, drivers, CUDA, networking (InfiniBand, RoCE), and virtualization
• Build automation for GPU provisioning, monitoring, and multi-tenancy
• Support AI/ML customers with high-performance GPU workloads

Skills

GPU Systems

HPC

AI Infrastructure

Kubernetes

Linux

Networking

CUDA

NVIDIA GPUs

Tools

NVIDIA GPU Operator

InfiniBand

RoCE

Job description

Confidential | AI & High-Performance Compute Platform. Our client is now powering the next wave of AI inference and real-time rendering workloads on their own GPU cloud.

Why This Role Is Exciting:

This is a GREENFIELD GPU cloud build. You will not just maintain GPUs, you will architect how thousands of GPUs are deployed, virtualized, and delivered as a service for AI customers. High demand, cutting-edge stack (H100, A100, L40S), zero legacy.

What You Will Own:

• Architect, deploy, and operate large-scale GPU infrastructure - bare metal + orchestration
• Optimize GPU performance, drivers, CUDA, networking (InfiniBand, RoCE), and virtualization
• Build automation for GPU provisioning, monitoring, and multi-tenancy
• Support AI / ML customers with high-performance GPU workloads

Who You Are:

• 4+ years with GPU Systems / HPC / AI Infrastructure
• Hands-on with NVIDIA GPUs, CUDA, driver stack, GPU operator, Kubernetes
• Strong Linux and networking fundamentals, experience with high-performance fabrics
• Bonus: Experience with GPU cloud providers, CoreWeave / Lambda / Vultr style platforms