hpc-ch forum on Business Continuity / Crises Management & Energy Savings and Efficiency - Hybrid Event

Thursday 23 May 2024, 08:00 → 16:15 Europe/Zurich

Christoph Witzig (Switch), Michele De Lorenzi (CSCS)

Description

In this event, we will discuss issues around professional data centre management, energy consumption/savings, how business continuity can be guaranteed, and what should be done in a crisis.

 

Key Questions

• How do you manage your data centre?
• Do you know the energy costs of your data centre? What are benchmarks? What savings are possible?
• How will energy consumption evolve in the future (AI, IoT, big data)? What are the constraints of your existing data centre? How do you scale?
• What needs to be done to guarantee the business continuity of your data centre?
• What do you do in case of crisis?

08:30 → 09:30 Facility Tour

We invite people living close to Lupfig to register for this slot. 
If you come from Zurich HB, you can catch the train at 07:36 and get to Lupfig at 08:18 (the venue is within 5 min. on foot). The hpc-ch forum will end at 15:00.

09:30 → 10:00 Welcome Coffee & Registration

10:00 → 10:15 Welcome and Introduction

Speakers: Christoph Witzig (Switch), Michele De Lorenzi (CSCS)

10:15 → 10:45 Data is the most Valuable Asset

Convener: Marco Stadler (Green)

10:45 → 11:15 Switch Cloud GPU Offering

Convener: Visho Jesudasan (Switch)

11:15 → 11:45 Disaster recovery for large scientific data: what are the options?

Ensuring academic research continuity in a major disaster is a challenge when dealing with data volumes in the petabyte range. In this presentation, we'll share our thoughts concerning disaster recovery at the University of Basel. We want to open the discussion and potentially seed synergies across organizations to address this challenge.

Speaker: Thierry Sengstag (University of Basel)

11:45 → 12:15 Decreasing the energy consumption of Euler

Speaker: Loic Hausammann (ETH Zurich)

12:15 → 13:15 Lunch and Networking

13:15 → 13:45 GitOps Built-In Disaster Recovery

Speaker: Nicolás Kowenski (ETH Zurich)

13:45 → 14:00 Integrating Operational and Energy Datasets in HPC

Speaker: Michele Brambilla (CSCS)

14:00 → 15:00 Community Development

• New community members
• Updates from the community

Efficient HPC Cluster Deployment in the Cloud: Strategies for Scalability, Speed, and Self-Service

This talk outlines strategies for deploying High-Performance Computing (HPC) clusters in the cloud to enhance operational efficiency and flexibility. We discuss our method for rapidly provisioning dedicated cloud clusters, providing a reliable overflow capacity during peak demands and a robust backup during on-premise outages or planned interventions. Our experience with seamless cloud bursting, alongside our future goals for user-driven resource management through AI-enhanced tools, highlights our commitment to advancing cloud-based HPC solutions. The presentation will detail the technical achievements and the roadmap for empowering users to autonomously manage computing and storage resources.

Speaker: Pablo Llopis, EPFL

15:00 → 16:00 Facility Tour

We invite people coming from Tessin and other Cantons to register for this slot.

16:00 → 16:10 Farewell and End of the Meeting