Lightning Talk: Does Cloud Elasticity Pay off for HPC Workloads?
Lightningat Kubernetes Batch + HPC Day Europe 2023 · on Tuesday, 2023-04-18· at 16:40· Hall 7, Room A | Ground Floor | Europe Complex·
summary
Ortec Finance moves HPC from on‑prem to managed OpenShift on AWS/Azure and assembles an open‑source stack that actually works: KEDA and Knative for bursty HTTP-to-job plumbing, Prometheus + Argo CD + ActiveMQ for ops, and templated job definitions. Queue-backed task runners reboot failed work and scale to zero. Empirical results: 1500 CPUs in ~20 minutes, Azure spot runs for ~20 minutes cost ~€6, while individual node provisioning averages ~4 minutes.
AI-written summary · about
chapters (7)
- 0:00Introduction
- 1:24Our journey so far
- 2:48Whats under the hood
- 4:15How it works
- 5:58How in a managed context
- 6:53Conclusion
- 8:39Where to learn more
related talks
34mHPC on AWS: Innovation without infrastructure constraints
18mKeynote: Mind the Gap: Bridging Cloud Native Innovation with Real-World Use
33mHPC on AWS: Innovation Without Infrastructure Constraints
47mCloud Native and Sustainable, Reproducible Scientific Computing
21mKeynote: CERN Experiences - Ricardo Rocha & Clenimar Filemon
35mHybrid Cloud Bursting Electronic Design Analysis Optical Proximity C....
16mA novel Cloud High Performance Computing infrastructure for user
27mFrom Supercomputing to Serving: A Case Study Delivering Cloud Native Foundation Models