Senior GPU Infrastructure Engineer, Full-Time, Onsite, Lahore

Blesso Cosmetics is seeking a highly skilled Senior GPU Infrastructure Engineer to lead the hands-on deployment and validation of an on-premises NVIDIA H200 GPU environment on HPE hardware. This full-time, onsite role in Lahore requires managing the entire process from physical installation to creating a validated, production-ready cluster. The engineer plays a critical part in ensuring that the GPU infrastructure meets performance and reliability standards before handing over to the AI engineering team for use-case deployment. nnThe ideal candidate possesses extensive experience in Linux system administration, GPU driver and CUDA stack configuration, along with mastery of networking technologies like InfiniBand and RoCE. This role requires working independently in a regulated enterprise environment, emphasizing disciplined documentation and precision in configuration and testing. While this position does not include managing a team, the engineer will engage with client systems and provide expert knowledge critical to maintaining high-performance computing infrastructure.
Responsibilities
Validate facility readiness including power, cooling, airflow, rack weight, and network connectivity to support the GPU infrastructure.
Perform rack and cable management for HPE H200 GPU nodes ensuring correct physical and network setup.
Update firmware including BIOS, iLO, GPU VBIOS, and NIC/HCA firmware to ensure hardware compatibility and stability.
Configure system settings such as IOMMU, large-BAR, and performance profiles optimized for GPU workloads.
Install and configure the operating system, NVIDIA data-center drivers, CUDA toolkit, Fabric Manager for NVSwitch, and GPUDirect/RDMA stack supporting InfiniBand or RoCE.
Deploy container runtime with NVIDIA Container Toolkit, set up Kubernetes clusters with GPU Operator, and configure Run:ai for GPU scheduling and fractional sharing.
Conduct comprehensive validation and burn-in testing including NVLink topology verification, DCGM diagnostics, NCCL multi-GPU/multi-node tests, and bandwidth and thermal/power stress tests.
Support initial deployment of NVIDIA Model Inference (NIM) servers and facilitate knowledge transfer through detailed documentation and runbooks.
Maintain disciplined documentation and clear communication with clients in a regulated environment to ensure compliance and operational excellence.
Troubleshoot kernel and driver issues related to CUDA and GPU hardware to maintain stable system performance.

Additional responsibilities may include collaborating with hardware vendors, updating systems for security and performance patches, and optimizing cluster configurations for emerging AI workloads.

Skills

Linux System Administration NVIDIA HGX High Performance Computing (HPC) Kubernetes Infiniband CUDA Firmware and BIOS Updating GPU Driver and Kernel Troubleshooting Container Runtime and NVIDIA Container Toolkit GPU Operator Deployment GPU Scheduling and Resource Sharing NVLink and NVSwitch Fabric Management GPUDirect and RDMA Networking Cluster Validation and Burn-in Testing Documentation and Runbook Creation

Job Details

Total Positions:

1 Post

Job Type:

Full Time/Permanent

Job Location:

Lahore, Pakistan

Gender:

No Preference

Minimum Experience:

5 Years

Apply Before:

Aug 29, 2026

Posting Date:

Jul 28, 2026

Blesso Cosmetics

· 301-600 employees - Lahore

Blesso Cosmetics have a diversified product portfolio which encompasses over 50 skin care products. Amongst them the most famous one are "WHITENING TUBES" that are directly imported from THAILAND. These tubes are being manufactured with keeping the Mission of "PREMIUM QUALITY PRODUCTS" ahead. Special blend of HALAL INGREDIENTS approved by islamic council make them one of a kind. BLESSO has also introduced quality products in different markets as per the demand of its prestigious customers as the company promises to place its user needs always on top priority. Our Manufacturing facilities are cGMP (Cosmetic Goods Manufacturing Practices) Compliant. Most of our products are 100% Halal Certified, and we adhere to strict regulations to assure Strength, Quality, and Purity of our Products.