Lightning Talk: Live Migration of PyTorch GPU Nodes From Azure To European Clouds - Mike Krom
About this talk
This talk covers the live migration of PyTorch GPU nodes from Azure to European cloud providers, addressing the sovereignty concerns for European teams using hyperscalers like Azure, AWS, or GCP for their ML workloads. The speaker, Mike Krom from Acf Cyber Solutions, demonstrates the process of migrating a GPU-enabled PyTorch virtual machine, including essential components such as CUDA drivers and Jupyter notebooks, without the need to retrain models or rebuild environments. Practical challenges like GPU compatibility, reproducibility, and data movement across clouds are discussed, with a focus on using DigitalNomadSky, an open-source platform for cross-cloud VM migration, illustrating how to maintain control over ML infrastructure while mitigating risks associated with jurisdiction and vendor lock-in.
More from this event
See all 103 talks →
What PyTorch Conference Europe 2026 Was Really Like – Official PyTorchCon EU Highlights | Paris
0:53
Lightning Talk: How DeepInverse Is Solving Imaging in Science and H... Andrew Wang & Minh Hai Nguyen
9:50
Why WideEP Inference Needs Data-Parallel-Aware Scheduling - Maroon Ayoub & Tyler Michael Smith
25:37
Write Once, Run Everywhere with Pytorch Transformers - Pedro Cuenca, Hugging Face
19:17