Use Jupyter Notebooks or the built-in Data Science Experience feature to quickly connect to system data and begin modeling immediately. Data scientists can access data regardless of location or data type. This allows them to maximize the value of their models by using the right data.
Embedded Spark processing with machine learning
Embeds Apache Spark processing to provide higher performance. Models can be deployed directly where the data resides, reducing data processing that would impact performance and increase complexity.
Common SQL engine
Shares a common analytics engine with other offerings in the IBM hybrid data management portfolio. The technology decouples the analytics application and data storage, enabling applications to work transparently with on-premise, cloud, RDBMS, NoSQL and Hadoop data sources. With IBM, you can easily move to or from the cloud.
Cloud-ready with scalable deployment
Scale up to petabyte levels. Flexible configurations allow you to expand computing and storage capacity independently, and scale your workload as needed. Features multi-temperature tiered storage. It ensures that highly accessed data resides on high-performance Flash, while older data resides on HDDs, to optimize performance and reduce cost. Workloads required to be in the cloud can also be moved without application rewrites.
High-speed architecture
Uses a massively parallel processing architecture to provide operational efficiency and low latency. Each node uses IBM Power processors and FlashSystem storage for advanced query performance. Also provides high-performance Db2 processing and BLU Acceleration for dynamic in-memory computing capabilities.
Easy to manage and maintain
Simplifies the management of your analytics system by reducing the maintenance required for tuning, indexing or aggregated tables.
You may also be interested in
IBM Lift
Easily perform ground to cloud database migrations