Hitachi Vantara has turned to Tel Aviv for a speed boost when it comes to crunching through massive data sets for analytics and delivering insight.
The Japanese data giant is partnering with SQream, and will offer the Israeli firm’s Data Acceleration Platform alongside its own object storage tooling.
The SQream product set will be integrated with both Hitachi’s Content Software for File NVMe parallel file system, and the Hitachi Content Platform object storage platform to accelerate analytics on Hadoop-Hive, Cloudera, Teradata and the like. It can also be turned towards running AI/ML models across massive data sets.
SQream’s technology consists of a database management system designed to work on – and be accelerated by – GPUs, whether on-prem, in the cloud or across both, and with datasets from “dozens of terabytes to petabytes”. The company claims ingest speeds of 3TB per hour per GPU.
At the core of SQream’s platform is a standard SQL database, with automatic turning, while the system is compatible with mainstream machine learning frameworks such as Spark MLib, R and Tensorflow.
The companies say they have already put the collaboration through its paces with a global manufacturer who used the combination for anomaly detection using AI and machine learning on an HPC platform to improve equipment efficiency. This involved continual analysis of a mutli-peta-scale database of machine sensor data across thousands of tables, they said.
SQream boosted its analytics service last year with its acquisition of Panopoly, with the aim of integrating the target firm’s no code ELT and data warehouse management tech into its own analytics service.
That buy followed its ingestion of $39.4m in series b+ funding in 2020. Its backers include Chinese tech giant, Alibaba Group.
