Announcing
eXtremeDB 9.0!
• Native ANN Indexes
• IoT Edge in 8 KB or Less
Small footprint data management for sensor data fusion
eXtremeDB offers unique features to optimize sensor data fusion.
Why consider small footprint eXtremeDB for sensor data fusion?
eXtremeDB’s small footprint and frugal use of memory and CPU uniquely qualify it to fuel database management for resource-constrained embedded and real-time applications. It was designed from the beginning for mission-critical systems, and protects against loss of database availability, safeguards data integrity, and is resistant to database corruption caused by application software defects.
NEW: eXtremeDB Edge Client
The eXtremeDB Edge Client is lightweight transactional storage for highly resource-constrained devices. It extends Active Replication Fabric to systems requiring as little as 8 KB of memory, allowing devices that cannot host a complete embedded database to participate directly in the distributed storage architecture.
The Edge Client supports multiple related data types and preserves referential integrity before replication. Even the smallest devices become full participants in the distributed storage architecture rather than simple message producers.
Uses the same schema, data model and transactional semantics as every other eXtremeDB edition, allowing the same application architecture to scale naturally from tiny edge devices to embedded controllers, gateways and enterprise servers.
The first and only commercial off-the-shelf (COTS) hard real-time database management system.
eXtremeDB/rt is the only COTS embedded database system to work with a RTOS to enforce database transaction deadlines, making it uniquely suitable for embedded systems with hard real-time constraints. It provides sophisticated development capabilities such as support for varied data and query types, native APIs, and a powerful debugging environment that includes a self-diagnostic API to catch costly defects before they slip into production code.
The pioneering eXtremeDB/rt design preserves the temporal validity of data and enforces predictable execution of critical database transactions, making it ideal for data management in hard real-time systems such as mission-critical sensor data fusion.
eXtremeDB supports shared data in Asymmetric Multiprocessing (AMP) configurations. AMP is ideal for sensor data fusion in that a single database can process both hard real-time data acquisition and AI/ML.
Asymmetric Multiprocessing (AMP)
Ideal for mixed-criticality systems
eXtremeDB/rt supports all major RTOS
Learn about data management in Advanced Driver Assistance (ADAS)
Active Replication Fabric™ mitigates connectivity issues.
eXtremeDB’s exclusive Active Replication Fabric™ mitigates IoT connectivity issues and compresses data, reducing storage space requirements by up to 75% and improving the speed of reading the database by up to 21%.
The Active Replication Fabric APIs allow device-based applications to collect data, then transmit the collected data when connected. Multi-tier means from an edge device to gateway to a server – as many hops as are needed according to the network topology. It also allows replication downstream from server to edge devices.
Data flow is fully automated through the Active Replication Fabric allowing developers to mitigate or even solve IoT connectivity issues.
eXtremeDB offers Run-length encoding (RLE) compression and low-level network compression.
Optimized for time series data.
Built on a core in-memory database system (IMDS) design with advanced features to deliver high performance, scalability and reliability, eXtremeDB offers the lowest possible latency for record-setting time series data processing.
A columnar data layout accelerates time series analysis, and a conventional layout using rows for data that is not sequential. eXtremeDB offers a flexible hybrid data design to optimize the performance of managing mixed data.
eXtremeDB is optimized with internal transaction and memory managers to provide maximum efficiency working with multi-threaded applications. Learn how.
Read the blog post, “How eXtremeDB ANN indexes and time series / sensor data fusion work together “.
A columnar data layout accelerates time series analysis, and a conventional layout using rows for data that is not sequential. eXtremeDB offers a flexible hybrid data design to optimize the performance of managing mixed data.
eXtremeDB is optimized with internal transaction and memory managers to provide maximum efficiency working with multi-threaded applications. Learn how.
How eXtremeDB ANN indexes and time series / sensor data fusion work together
In simple terms: sensor data → eXtremeDB time-series processing → vector → ANN search → matching patternsBoth the time-series data and ANN indexes stay in eXtremeDB and can run locally on the device. Specialized signal processing or an external ML model can be added where needed, without moving the data to a separate specialized vector database engine or becoming dependant on the cloud.
Learn about eXtremeDB database indexes
Features to optimize time series data
Related resources
White Papers for Professional Developers
Read Shared Data in Asymmetric Multiprocessing (AMP) Configurations.
Review our other research
Webinars for Professional Developers
We invite you to watch the many on-demand Webinars hosted by McObject database experts, such as Database System Requirements to Achieve Level 3+ Autonomous Driving or Embedded Databases: Building In Always On High Availability
Review our list of Webinars
Articles for Professional Developers
- Dogmatic views on safe languages used for mission- and safety-critical applications. A blog highlighting a few challenges and sharing a few lessons learned about memory management in embedded applications for safety-critical system development, reliability and dependability. Find the full text at Embedded.com.
- Buy vs. Build for Time-Critical Sensor Data Fusion. What exactly is “real-time”? This article looks at its permutations and offers insight into time-critical sensor data fusion and its role in database systems. Read the article in ElectronicDesign.com
See a list of articles