IBM is expected to reveal its next generation grid-based XIV Storage System this week, which they hope will appeal to performance-conscious customers with virtualized server environments, big data analytics operations and cloud computing setups.

The new generation of XIV won’t arrive for customers ready to fork over what ComputerWorld reported as $2 million until September, but news about the improvements, especially for the growing numbers of virtualized, big data and cloud customers are looking for enhanced options.

According to a report in advance of the news, IBM will tout the XIV Storage System as being four times faster than the last model with enhanced features that will ease management and allow the system to support more workloads, making it a better fit for enterprise users.

As Lucas Mearian stated in ComputerWorld today, “with this release, IBM moves from Intel Nehalem processors to the latest Westmere chips. It also upgraded from a gigabit Ethernet backbone to an InfiniBand interconnect and moved from 4 GB/sec Fibre Channel to 8 Gbit/sec Fibre Channel front-end ports.”

Each of the new generation models will come standard with two InfiniBand switches with redundant inter-module connectivity for up to 600Gbit/sec total internal bandwidth and an increase in the number of iSCSI ports—going from a mere six to 22.

IBM’s VP of enterprise disk storage at IBM told ComputerWorld that “we’re starting to see demand pick up for IP connectivity, though I’d still say it’s slower than what we in the industry predicted it would be by this point in time…this will help customers get prepared for that future transition into greater workload for IP connectivity.”

In addition, Cancilla also said that IBM plans to put an SSD into every disk drive drawer to act as a caching layer, sitting between the controller and the spinning disk. This will have all the performance benefits of SSD but it doesn’t complicate management from a data tiering standpoint.”

According to Cancilla, the grid architecture of the XIV line as a whole allows performance to grow with the addition of new disks, saying that the simple configuration of the array was planned so that set up and management wouldn’t be a hassle.

Mearian added that while the array still doesn’t migrate data across multiple disk types depending on performance needs, it is able to scale up from 27 terabytes of capacity to 161 terabytes. Furthermore, he adds, IBM has added the ability to perform non-disruptive code updates, data snapshot possibilities as well as synchronous and asynchronous replication."

Full story at ComputerWorld