Skip to content
Search jobs inside your AI agent. Claude, ChatGPT, Claude Code, Codex, Gemini, Cursor and Grok
Connect

SVI Public Company Limited

NewBe an early applicant

Full Stack Software Engineer – Data center applications

Posted

Description

Role Overview

We are looking for a Lead Systems Software Engineer (GPU Infrastructure) to architect, deploy, and scale the software platform powering our newly established GPU data center cluster. You will bridge the gap between bare-metal accelerated hardware and the software stack required to run massive AI training, inference, or high-performance computing (HPC) workloads.

Your primary responsibility will be creating a highly resilient, automated, and observable software layer that maximizes GPU utilization, eliminates networking bottlenecks, and offers seamless scheduling capabilities to internal or external development teams.

Core Responsibilities

  • Cluster Orchestration & Scheduling: Design, implement, and maintain multi-node GPU clustering software using orchestration frameworks like Kubernetes (with advanced scheduling plugins like Volcano, Kueue) or Slurm to handle heavy batch processing workloads.
  • Hardware-to-Software Optimization: Write, configure, and optimize low-level drivers, runtimes, and libraries (e.g., NVIDIA CUDA, NCCL, Triton Inference Server) to ensure deep learning frameworks can communicate across the cluster with zero latency.
  • Performance Bottleneck Analysis: Identify and eliminate systemic performance drops across host CPUs, PCIe buses, high-speed interconnects (InfiniBand/RoCE), and high-performance storage systems (e.g., Lustre, GPFS, Weka).
  • Infrastructure Automation (IaC): Build robust automation scripts and configuration management tools using Ansible, Terraform, and Helm to provision, patch, and deploy bare-metal server fleets dynamically.
  • Telemetry & Observability: Develop and maintain specialized telemetry systems (using Prometheus, Grafana, and DCIM integrations) to monitor GPU temperatures, power consumption, memory errors (XIDs), and structural utilization metrics.
  • Collaboration: Act as the technical translator between the data center facility engineering team (managing power and liquid cooling) and the product engineering teams (running the AI models).

Required Technical Skills & Qualifications

  • Experience: 5+ years in Systems Engineering, Infrastructure Engineering, or low-level DevOps, with at least 2 years explicitly handling production-scale GPU or HPC workloads.
  • Programming: Strong proficiency in Python, Go, or C++ for writing system tooling, automation, or custom infrastructure components.
  • Linux & Networking: Master-level knowledge of production Linux (Ubuntu/RedHat), networking internals (TCP/IP, routing, firewalls), and advanced cluster networking like InfiniBand or RoCE (RDMA over Converged Ethernet).
  • Containerization Stack: Deep experience managing production Kubernetes environments, Container Runtime Interfaces (CRI), and NVIDIA Container Toolkit integrations.
  • Hardware Literacy: Familiarity with modern data center GPU architecture (e.g., NVIDIA H100/B200 platforms), high-bandwidth memory (HBM), and interconnect systems like NVLink.

Preferred / "Nice to Have" Qualifications

  • Experience implementing liquid-cooling software alerts or digital twin facility
  • telemetry.
  • Contributions to open-source distributed training frameworks or cloud-native batch scheduling projects.
  • Certified Kubernetes Administrator (CKA) or formal NVIDIA/HPC systems architecture certifications



Skills

Apply

See also

Full-Stack jobs by country — openings, pay and top skills →