Azure Databricks | Microsoft Azure (original) (raw)

• Azure Databricks is a fast, easy, and collaborative Apache Spark-based data and AI platform optimized for Microsoft Azure. It provides a unified environment for big data and AI workloads, combining the best of Databricks and Azure to simplify data engineering, data science, and machine learning.
• No, Azure Databricks is not a database or a storage system. It’s a data analytics and AI platform that enables users to process, analyze, and visualize large volumes of data and AI use cases. Azure Databricks provides database-like capabilities on top of cloud object storage and integrates with various data sources, including databases, data lakes, and cloud storage.
• Unity Catalog is a unified governance solution for all data and AI assets in Azure Databricks. It provides centralized access control, auditing, lineage, and data discovery across workspaces. Unity Catalog simplifies data governance by enabling fine-grained access policies and ensuring consistent security and compliance across your data estate.
• A Databricks unit, or DBU, is a normalized unit of processing capability per hour based on Azure VM type and is billed on per-second usage. The DBU consumption depends on the size and type of instance running Azure Databricks.
• Serverless compute in Azure Databricks helps you run workloads without managing infrastructure. It scales automatically and is fully managed to enable fast startup and simplified operations.
• Photon is a high-performance, vectorized query engine built in C++ that accelerates SQL and DataFrame workloads in Azure Databricks. It improves speed and lowers costs without requiring code changes.
• The default format for all data tables, Delta Lake is an open-source storage layer in Azure Databricks that brings atomicity, consistency, isolation, durability (ACID) transactions, scalable metadata, and unified batch and streaming data processing to your lakehouse.
• You can save on your Azure Databricks unit (DBU) costs when you pre-purchase Azure Databricks commit units (DBCUs) for one or three years. You can use the pre-purchased DBCUs at any time during the purchase term.
  The pre-purchase discount applies only to the DBU usage. Other charges such as compute, storage, and networking are charged separately.