GPU Infra Management

Role Title

Skills Required

Roles and Responsibilities

GPU Infra Management Expert

 . GPU System Engineer to manage high-performance GPU systems and develop software solutions to optimize GPU resource utilization for scheduling-based applications.

. Well versed with Nvidia Legacy & New GPU systems (RTX, v100, A100 etc) with MLOPs, Orchestration, Kubernetes, Slurm, or similar systems  is must.

 1. GPU System Management:

- Deploy, configure, and maintain GPU hardware and systems.

- Monitor GPU resource usage, health, and performance for various workloads.

- Manage GPU clusters to ensure optimal availability, stability, and scalability.

-  MLOps GPUs infra setup and management

2. Software Development:

- Design and develop GPU management software to automate scheduling and resource allocation.

- Build tools and scripts for GPU workload optimization, resource monitoring, and job scheduling.

- Develop effective GPU scheduling algorithms for multi-application environments.

- Integrate scheduling tools with containerized or virtualized environments like Kubernetes, Slurm, or similar systems.

3. Software Deployment:

- GPU Orchestration S/W deployment

- Configuration Setup and update

4. Performance Optimization:

- Analyze GPU performance metrics and fine-tune systems for improved efficiency.

- Optimize GPU utilization for AI/ML, rendering, and high-performance computing workloads.

5. Troubleshooting and Support:

- Diagnose and resolve GPU hardware/software issues.

6. Documentation and Reporting:

- Document GPU architecture, processes, and management tools.

- Provide detailed reports on GPU utilization, performance trends, and improvements.