DevOps Engineer, GPUaaS

Singtel Group; Singapore; On-site; SGD 120,000 - 170,000; Full time

Job summary

Singtel Group is seeking a DevOps Engineer for its GPUaaS platform to design, deploy, and operate large-scale GPU clusters for AI/ML workloads across on-prem and cloud environments.

You will implement CI/CD pipelines, automate provisioning, monitor health and performance, and optimize GPU/HPC infrastructure, collaborating with software and admin teams.

Qualifications

• Bachelor's degree in Computer Science, Engineering or related field.

Responsibilities

• Design, deploy and support large-scale GPU clusters for AI/ML workloads.

Skills

Jenkins

Kubernetes

Ansible

Terraform

Python

Bash

Prometheus

Zabbix

TensorFlow

PyTorch

MPI

NCCL

Docker

InfiniBand

RoCE

GPU/AI

Education

Bachelor's in CS/Engineering

Tools

Docker

Kubernetes

Slurm

Terraform

Ansible

Prometheus

Zabbix

Job description

Select how often (in days) to receive an alert:

Singtel Digital InfraCo’s RE:AI division is building Asia’s most advanced and sustainable AI infrastructure ecosystem. RE:AI enables enterprises, research institutions, and digital-native businesses to accelerate innovation through responsible, high-performance AI compute and connectivity solutions.

Be a Part of Something BIG!

As an DevOps Engineer for SingTel’s GPU-as-a-Service (GPUaaS), you will help in implementing processes and integration of operations to advance customer’s AI and HPC capabilities. You will be exposed to both physical data center implementation and software solutions in a Singtel GPU-as-a-Service (GPUaaS). This position requires a forward-thinking individual who thrives in dynamic environments and is committed to driving continuous improvement in GPU for AI and HPC environments. This role is suitable for professionals looking to develop their expertise in DevOps and AI/HPC cloud platforms.

Make an impact by

• Design, deploy and support large-scale, distribute GPU clusters for AI and ML workloads.
• Manage and automate provisioning of GPU resources in both on-prem and cloud platforms.
• Design, implement and manage CI/CD pipelines for AI models and GPU-accelerated applications.
• Monitor cluster usage, health, performance and availability.
• Improve infrastructure provisioning, management, and monitoring through automation.
• Troubleshoot compute resource system level issues such as Slurm, Kubernetes, GPU drivers, CUDA, IB networking.
• Optimize system parameters (e.g., OS, drivers, networking, library) for AI workload performance.
• Conduct GPU cluster benchmark and keeping up with the latest advancements in GPU technology.
• Set up monitoring and logging for GPU resources using Zabbix, Prometheus, NVIDIA DCGM and other tools.
• Implement security best-practices for multi-tenant GPU-as-a-Service (GPUaaS) environment.
• Collaborate with software and administrator to to streamline workflows and improve collaboration.
• Providing technical support and guidance to users of GPU-accelerated systems.
• Work with senior DevOps engineer to identify bottlenecks and improve development and operational processes for AI and HPC GPU cloud.
• Learning to solve problems in high-performance distributed computation for AI and HPC GPU cloud computing.
• Participate in rotational or scheduled shift work as required to support platform operations.

Skills for success

• Bachelor’s degree in Computer Science/Engineering, Information Technology, Systems Engineering, or a related field.
• Experience with DevOps tools such as Jenkins, Kubernetes, Ansible and Terraform.
• Solid understanding of DevOps practices, including CI/CD, automation, and monitoring.
• Proficiency in scripting languages (e.g., Python, Bash).
• Experience in implementing monitoring solutions such as Zabbix, Prometheus.
• Familiarity with AI frameworks such as TensorFlow, PyTorch.
• Understanding of cloud architectures (IaaS, PaaS), GPU architecture and NVIDIA GPUs.
• Strong verbal, written, and presentation skills in English.
• Team player with experience in cross-functional coordination.
• Strong technical problem solving and analytical skills for system optimization.
• Understanding of how collective communications (MPI, RDMA, and NCCL) works, as well as an understanding of GPU specific acceleration works on GPU cluster.
• Knowledge of DevOps/ML Ops technologies in GPU cluster such as Docker/containers, Kubernetes, data center deployments
• Familiarity with Slurm or other HPC workload managers to manage GPU clusters.
• Understanding of AI & HPC networking technologies such as InfiniBand, RoCE, DPUs.
• System-level experience specifically GPU-based systems (NVIDIA GPU and SDKs)
• Understanding how AI and HPC workloads interact with both GPU HW and SW infrastructure.