Below are high level feature comparison between SQL Server 2016, 2017 and 2019. For more details please refer to Microsoft page.

Feature Availability:

 Not supported  Fully supported

Features	SQL Server 2019	SQL Server 2017	SQL Server 2016
Intelligence across all your data
BigData Clusters Data virtualization using PolyBase[1]
Database performance
In-memory database: in-memory OLTP
In-Memory Database: Persistent Memory[2]
In-memory database: memory-optimized tempdb
Real-time operational analytics
Intelligent query processing
Intelligent performance
Availability
Always on
Accelerated database recovery
Read scale-out availability groups
Security
Always encrypted
Always encrypted with secure enclaves
Data discovery and classification
Transparent data encryption
Backup encryption support
Encryption at rest and in motion
Dynamic data masking and row-level security
Cloud readiness
Backup to Azure
Disaster recovery to Azure
Optimized virtual machine (VM) images in Azure gallery
Free asynchronous replication on Azure Virtual Machines for disaster recovery
Management and programmability
Linux support
Container support
Kubernetes support
Temporal tables
JSON support
Graph data support
UTF 8 support
Java support
Azure Data Studio to manage SQL Server including support for T-SQL using notebooks
Database compatibility certification
BI and Analytics Support
Integration services managed as a server
Tabular BI semantic model[3]
Master data services[3]
Data quality services[3]
Many-to-many relationships in tabular models
End-to-end mobile BI on any device
SQL Server Analysis Services (SSAS) direct query
Calculation groups in tabular model

^{[1] For HDFS and Azure Blob Storage only(For SQL Server 2017, SQL Server 2016 only).}

^{[2] For tail of the log only (For SQL Server 2017, SQL Server 2016 only).}

^{[3] SQL Server Enterprise edition only.}