Glossary

Massively parallel processing

An architecture that splits a query across many independent compute nodes, each processing its own data slice.

Also called: MPP, MPP database

Massively parallel processing (MPP) is a database architecture in which a query is broken up and executed simultaneously across many independent compute nodes, each with its own CPU, memory, and typically its own slice of the data, coordinated by a leader node that distributes the work and assembles the final result. It is the architecture behind most modern cloud data warehouses.

This differs from a single, powerful server scaling up, which eventually hits a hardware ceiling; MPP instead scales out by adding more nodes, and query time drops roughly in proportion to how well the data and work can be divided across them. How rows are distributed across nodes — often driven by partitioning or a hash on a key column — directly affects performance: a poor distribution key leaves some nodes idle while others do most of the work, a problem known as data skew.

MPP is what allows a warehouse to scan and aggregate billions of rows in seconds, and it pairs naturally with columnar storage, since both techniques reduce the amount of data any one node has to read. The main operational pitfall is skewed data distribution or a join on a low-cardinality key, either of which can leave the overall query waiting on one overloaded node no matter how many nodes are available.

Last reviewed September 22, 2026

In the index now

Related terms

Query engine The software component that parses, plans, and executes queries against one or more data sources. Columnar storage A file and database layout that stores each column's values together, instead of storing each row together. Partitioning Physically splitting a table's data into segments, usually by a column like date, so queries can skip irrelevant ones. Distributed computing Splitting a computation across multiple machines that coordinate over a network to solve it faster or at larger scale. Data warehouse A centralized, structured database optimized for fast historical reporting rather than day-to-day transactions. Apache Spark An open-source distributed computing engine for processing large datasets in parallel across a cluster of machines.

Related tools

Amazon Redshift AWS's managed cloud data warehouse, available as provisioned clusters or an auto-scaling serverless option. Apache Doris Open-source MPP database combining real-time analytics with sub-second query response at large scale. Exasol In-memory, massively parallel analytics database available as a managed cloud service or self-hosted software. Greenplum Open-source, massively parallel PostgreSQL-based data warehouse for large-scale analytical SQL. StarRocks Open-source MPP database for sub-second analytics across real-time data and data-lake tables. Teradata VantageCloud Teradata's cloud-delivered evolution of its enterprise MPP data warehouse, unified under a single Teradata Unit pricing currency.

Related guides

How to choose a cloud data warehouse Start from your cloud, your pricing tolerance and your workload shape — general BI, high-concurrency apps, or enterprise mixed workloads — not the feature list. How to choose a real-time OLAP database Real-time OLAP databases trade operational complexity for speed on fresh data — the decisions are architecture, query concurrency, and who runs it.