Here’s a great collection of Jupyter notebooks that explore all the new features of SQL Server 2019.
Here are some of the ones that caught my attention.
SQL Server 2019 Querying 1 TRILLION rows
- OneTrillionRowsWarm.ipynb – This notebook shows how SQL Server 2019 reads 9 BILLION rows/second using a 1 trillion row table using a warm cache,
- OneTrillionRowsCold.ipynb – This notebook shows how SQL Server 2019 performs IO at ~24GB/s using a 1 trillion row table with a cold cache.
Big Data, Machine Learning & Data Virtualization
- SQL Server Big Data Clusters – Part of our Ground to Cloud workshop. In this lab, you will use notebooks to experiment with SQL Server Big Data Clusters (BDC), and learn how you can use it to implement large-scale data processing and machine learning.
- Data Virtualization using PolyBase – The notebooks in this SQL Server 2019 workshop cover how to use SQL Server as a hub for data virtualization for sources like Oracle, SAP HANA, Azure CosmosDB, SQL Server and Azure SQL Database.
- Spark with Big Data Clusters – The notebooks in this folder cover the following scenarios:
- Data Loading – Transforming CSV to Parquet
- Data Transfer – Spark to SQL using Spark JDBC connector
- Data Transfer – Spark to SQL using MSSQL Spark connector
- Configure – Configure a spark session using a notebook
- Install – Install 3rd party packages
- Restful-Access – Access Spark in BDC via restful Livy APIs
- Machine Learning
- Powerplant Output Prediction – This sample uses the automated machine learning capabilities of the third party H2O package running in Spark in a SQL Server 2019 Big Data Cluster to build a machine learning model that predicts powerplant output.
- TensorFlow on GPUs in SQL Server 2019 big data cluster – The notebooks in this directory illustrate fitting TensorFlow image classification models using GPU acceleration.