Actian Vector for Cloud
High-performance vectorized columnar analytic database on your choice of cloud
Actian Vector in the Cloud
Actian Vector, a columnar in-memory relational database engine designed for high performance analytics, is available on both Amazon Web Services and Microsoft Azure using the bring-your-own-license (BYOL) model, for both single-node and cluster configurations. The Actian Vector Community Edition is available as an Amazon Machine Image (AMI) for 1-click deployment through the AWS Marketplace, and also on the Microsoft Azure Marketplace. Deploy the same fast analytics both on premise and in the cloud.
|Deployment Models:||Vector Enterprise Edition||VectorH Enterprise Edition||Vector Evaluation Edition||Vector Community Edition|
|BYOL for Amazon AWS|
|BYOL for Microsoft Azure|
|AMI for AWS Marketplace|
|Microsoft Azure Marketplace|
High-Performance Vectorized Columnar In-Memory Analytic Database
Built for Speed
Actian Vector is designed for speed and efficiency using column-based storage and vector processing to deliver record-breaking in-chip analytics.
Built for Open
Actian Vector enables broad access using open standards and provides extensibility through open source technologies like Spark and Hadoop.
Built for the Enterprise
Actian Vector delivers a unique combination of cutting edge innovation and mature database features that are proven in the enterprise.
Vector performance advantages extend to cloud-based configurations as well, based on testing recently done by MCG, an independent global services firm specialized in information strategy and implementation services. Vector’s performance advantage over Amazon Redshift, MS SQL Server, Snowflake, and Cloudera Impala increases as database size, query complexity, and user concurrency increases.
More than 20X Faster than Snowflake
Against Snowflake, MCG showed that Actian Vector is on average 6-12 times faster on a sixteen-node AWS configuration at 1-, 5-, and 10TBs, and up to 20 times on join queries. With 20 users, Vector demonstrated significantly faster times than Snowflake, even beating Snowflake by 17X with its multi-cluster autoscaling feature turned on (at 8X the cost) to accommodate the additional users.
Up to 14X Faster than Amazon Redshift
Based on testing using the Berkeley AMPLab big data benchmark, Vector is consistently 2-3 times faster than Redshift on a 5-node cluster, scaling from 1- to 5- to 10TBs of data. That advantage for Vector increases with the capacity and complexity of the scan, aggregation and join workload, to nearly 14 times on the 10TB join query with 20 concurrent users.
Almost 10X Faster than MS SQL Server
Using the same Berkeley AMPLab benchmark, MCG showed that Actian Vector is 3-6 times faster than Microsoft SQL server on a single-node AWS configuration at 500GBs, 750GBs, and 1TB. On concurrency testing with 20 users, Vector scaled up to nearly 10 times faster than Microsoft, which couldn’t even complete some of the tests at 1TB capacity.
Over 100X Faster than Cloudera Impala
The Berkeley AMPLab benchmark was originally created to measure Impala against other big data query engines. MCG benchmarks show Actian Vector nearly 40 times faster on average than Cloudera Impala on a 16-node AWS cluster with 1, 5, and 10TBs. Impala failed to complete the largest join query at the 10TB capacity for a single user, and most of the tests for 20 users.
Data-driven organizations rely on analytic databases to load, store, and analyze volumes of data at high speed to derive timely insights. This benchmark addresses some fundamental business questions that any organization might encounter and ask. Actian Vector, through this industry standard set of scan, join and aggregation queries, demonstrated a significant performance advantage over Amazon Redshift and Microsoft SQL Server. These objective results were driven by the patented Actian X100 vectorized query engine which exploits the parallelism capabilities of modern server hardware.
– William McKnight, President, MCG Global Services
Flexible adaptive parallel execution algorithms to maximize concurrency while enabling load prioritization