Infinidat customers can easily consolidate a number of corporate databases into a single InfiniBox system, reducing the occupied area and cost of network ports tenfold. InfiniBox prevents the uncontrolled growth of arrays and file systems by allowing exploitation and resource planning teams to take advantage of the scale effect created by unified storage pools whose capacity is comparable to that of a data center. Flexible software quality and mainframe reliability ensure that key business applications run seamlessly in parallel with applications used for development and testing in a secure, scalable, and efficient storage infrastructure.
The InfiniBox enterprise-class storage system outperforms all-flash performance, provides high availability when working with mixed workloads and allows you to increase the capacity to several petabytes, and data replication on the “active / active” scheme significantly increases business flexibility.
InfiniBox is based on Neural Cache – a machine learning algorithm that takes into account the historical relationships between sections of data received by the system from the beginning of use. Observing all network traffic going to the storage system, Neural Cache analyzes the history of I / O operations, finds correlations in data access patterns and immediately begins to predict future requests, leaving them in the cache. Because most I / O requests are handled directly from RAM (where latency is more than 100 times less than in flash memory), any application on InfiniBox will run faster than on traditional all-flash arrays.
Access to the petabyte data set in milliseconds radically accelerates the digital transformation in any field, be it genetic engineering, precision medicine, training drones and creating virtual reality games, banking or fraud detection.
