Apertus Engineer: Infrastructure

Seeking a Skilled Infrastructure Engineer

We are in search of a dedicated and proficient engineer to play a crucial role in managing the container image stack for a pioneering project. This individual will collaborate with experts to ensure high-performance infrastructure for large-scale training workloads.

Project Background

The initiative is driven by a coalition of leading institutions and aims to enhance AI capabilities through the training of extensive foundational models. The successful candidate will focus on maintaining the container image stack, ensuring stability and efficiency in close cooperation with our engineering team.

This impactful project strives to train models on one of Europe's largest AI-ready supercomputers, and involves a close-knit team of engineers and researchers focused on delivering open-source, responsibly trained AI solutions.

Key Responsibilities

• ML System Image Maintenance:
  • Develop and manage container images for all core ML phases, ensuring they are stable and efficient.
  • Handle the complete dependency stack for the ARM-based architecture and ensure reproducibility of image builds.
  • Collaborate with team members to validate and document images against prescribed workloads.
• Compute Partnership and Efficiency:
  • Act as the primary point of contact with engineers and researchers for compute-related inquiries.
  • Work on improving performance and efficiency of underlying compute infrastructure.
  • Document best practices and institutional knowledge regarding HPC systems.
• Infrastructure Stress Testing:
  • Conduct stress tests on infrastructure using representative workloads to validate system performance.
  • Collaborate with engineers to resolve any issues affecting stability and throughput.
  • Support and maintain the serving stack dependent on the same images.

Candidate Profile

Essential Qualifications:

• MSc or PhD in a relevant technical field; exceptional candidates may have notable experience without a graduate degree.
• Hands-on experience in HPC environments and strong proficiency in Linux containers.
• Excellent collaboration and communication skills, with a proven ability to engage across diverse teams.
• Flexibility to adapt to shifting priorities and tools in a dynamic environment.

Preferred Qualifications:

• Familiarity with LLM training frameworks and experience with ARM64 platform containers.
• Background in HPC networking and parallel filesystems.
• Experience with CI/CD pipelines for container image development.

Nice to Have:

• Publications in relevant fields or knowledge of recent research.
• Experience with profiling distributed workloads and contributing to open-source projects.

Workplace and Offerings

Joining this initiative provides an opportunity to work within a prestigious academic setting, alongside leading experts and researchers. The benefits include attractive employment conditions, professional development opportunities, and flexible working arrangements to suit different needs.

Diversity and Sustainability Commitment

We are dedicated to fostering an inclusive and diverse workplace, promoting equality, sustainability, and a respect for the rights and dignity of all staff and students, aligning with our core values.

Enthusiastic About Joining Us?

If you are intrigued by the possibility, we encourage you to explore further details and discover how you can contribute to groundbreaking advancements in AI technologies.