Secure data processing with AMD Confidential VMs on Azure Databricks
summary
Microsoft Databricks runs on AMD SEV‑SNP confidential VMs (example DC8as_v5) to encrypt data in use while storing encrypted shards on Azure Data Lake Gen2; demo shows AES UDFs that insert and later decrypt 5 encrypted records (names decrypted, other columns remain encrypted), requires VNET injection, private link, and managed disk encryption via managed HSM, and a customer (Telefonica Germany) reports production adoption and successful protection of sensitive analytics workloads.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:17Regulatory challenges
- 2:40Confidential computing
- 5:14Business value
- 7:42Reference architecture
- 9:50Live demo
- 13:23Customer testimonial
- 14:06Conclusion
related talks
12mUsing AMD based confidential VMs for your Azure Data Explorer clusters
22mSecuring the Future with Azure Confidential Computing
48mProtecting the confidentiality and integrity of your high value data with Azure SQL and SQL Server 2
46mAMD + Azure: Outstanding performance and the latest security features
13mVM Security with Confidential Computing
23mAMD SEV & Azure Local: When Confidential VMs Go Local
43mConfidential development with Always Encrypted using enclaves
43mAnalyzing Big Data in Azure cloud using Spark and Databricks