ULTIMATE PERFORMANCE
Real-time ingestion with throughput as high as hundreds of millions of records per second. Query PBs of data in sub-second latency.
LOWER COST
Disaggregated compute and storage, scale resources up and down according to your business needs. Unified realtime and offline storage.
HIGHLY AVAILABLE AND RELIABLE
Data is replicated in multiple copies by default, keep your data service always on.
Multi Cloud: Deploy on Kubernetes
Massively Parallel Processing (MPP) of SQL queries with C++ native engine enables high resource efficiency, and results in sub second latency.
POSTGRESQL COMPATIBLE, EASY TO USE
Compatible with PostgreSQL 11. Easily connect to BI and ETL tools (like Tableau, QuickBI, etc) via JDBC.
DIVERSE DATA ACCESS PATTERN
Supports row and column storage, meets your diverse query requirements such as: Point-Lookup Query (Hundreds of millions per second), Complex Analytical Query (Massive parallel processing), and Ad-hoc Query (No data migration).
- High ThroughputIngest real-time and batch data at extremely high throughput
(hundreds of millions of RPS)
Data ingested is immediately queryable - Compatible with PostgreSQLReuse all existing PostgreSQL tools without any migration
Leverage existing rich PostgreSQL extensions - Fast Point-LookupExtremely fast point-lookup
- Low LatencyQuery PB-scale data in sub-second latency
Maximizing business values with multi-dimensional interactive analytics - BI NativeSeamlessly connect to your favourite BI tools, like Tableau, PowerBI, Qlik, AWS Quicksight, via JDBC/ODBC
- High ConcurrencyRun hundreds of large queries at the same time without compromising query performance or interfere with each other
- Schema EnforcementValidate your input data for free
Get no surprise when retrieving and using your data - Built-in CDC (alpha)Ultimate simplicity
Replicate your production OLTP database to Hologres directly for OLAP without any middleware required - Upsert At-Scale As You WishUpdate and delete data at-scale with strong consistency
- Flexible DeploymentCloud-Native
Minimum deployment dependency
Run anywhere: single machine, bare metal, Docker, Kubernetes - Horizontally ScalableNo performance bottlenecks, scales linearly, single instance over 10,000 cores
- Battle TestedWhile handling 130 million real-time data records per second during Alibaba 11.11 Global Shopping Festival in 2019, Hologres returned 99.99% of queries within 130ms. Single largest table over 2.5PB

Data Storage and Data Services