Micro Focus Announces Vertica 10, Delivering Unified Predictive Analytics at Massive Scale

By AIT News Desk On Mar 31, 2020

Leading Analytical Data Warehouse Empowers Organizations to Derive Greater Insight at Extreme Scale by Unifying Data Siloes Across Cloud and Hybrid Environments; Features Include Advanced Integration with TensorFlow and Python to Operationalize Machine Learning at Scale

Micro Focus announced the Vertica 10 Analytics Platform, which includes major updates for operationalizing machine learning at scale and expanding deployment options for Vertica in Eon Mode, enabling the most intensive variable workloads across major cloud and on-premises data centers. With Vertica 10, organizations are better equipped to unify their data siloes and take advantage of the deployment models that make sense now and in the future in order to monetize exponential data growth and capture real-time business opportunities.

“Over the years, many organizations have successfully captured massive amounts of data, but are now challenged with getting the business insights they need to become data-driven. The market demand to leverage cloud architectures separating compute from storage needs to be balanced with the higher costs and increased risk of cloud-only data warehouses, while machine learning projects with tremendous potential have struggled to make their way into production,” said Colin Mahony, Senior Vice President and General Manager, Vertica, Micro Focus. “Vertica 10 expands the options for a unified analytics strategy to address growing data siloes, a mix of cloud, on-premises, and hybrid environments, and the pressing need to operationalize machine learning at scale.”

Vertica 10 offers deeper integration with Python and TensorFlow for unsupervised learning and PMML standard model format for cross-platform compatibility. By offering integration with these popular languages and tools, data scientists can continue using Python and TensorFlow, while leveraging larger volumes of data and parallelized performance advantages to improve accuracy and replicability.

Vertica 10 expands deployment and communal storage options for Vertica in Eon Mode. With Vertica in Eon Mode’s expanded public cloud support for Google Cloud Platform (GCP) and Apache Hadoop HDFS as communal storage, organizations now have more options to optimize infrastructure costs and simplify operations. Following support for Vertica in Eon Mode for Pure Storage FlashBlade, updates to Vertica in Eon Mode for HDFS and MinIO now include more choices for organizations to manage their dynamic workloads with S3 object stores in on-premises and cloud environments. The expansion of Vertica in Eon Mode, already available on AWS and now on GCP, expands the public cloud options for managing analytically intensive, dynamic workloads.

Highlights and enhancements to Vertica 10 also include:

Seamless migration – New migration functionality enables a seamless migration for customers moving from Vertica in Enterprise Mode to Vertica in Eon Mode to adopt the next-generation data architecture separating compute from storage for on-premises, hybrid, and cloud deployments.
Next-generation Vertica Database Designer – The updated Vertica Database Designer offers significant improvements to operations and ease of use, reduces resource usage by orders of magnitude, and improves projection designs for consistently faster queries.
Range of security improvements – Vertica 10 includes security enhancements that span simplifying the process of administering TLS certificates, better user authentication and permissions management with LDAP Link, Kerberos for vertica-python and new permissions system tables, and improved support for format preserving encryption with new Voltage integration capabilities.
Operationalizing machine learning at scale with PMML and TensorFlow integration – Vertica 10 now imports models built in other platforms and languages like Spark, Python, and SPSS using the PMML format. With PMML model export, models built in Vertica can also be exported for scoring in other systems such as edge nodes for IoT use cases. Organizations can also now put Neural Networks and custom machine learning models into production by importing pre-trained TensorFlow models into Vertica for predictions on hot data and archiving for replicability.
Analyze data in place and in any format – With Vertica 10, organizations can now analyze complex data types from Maps and Arrays to Structs in Parquet on S3 or HDFS to open SQL-based analytics to new use cases.