Hitachi Vantara New VSP E-series and HNAS systems

In Hitachi NAS, Hitachi Vantara, Hitachi VSP by AdministratorLeave a Comment

Hitachi Vantara recently announced new additions to their midrange, VSP E Series systems, the E590 and E790 as well as updates to their HNAS platform with new 5200 and 5300 systems.

Hitachi VSP E590 and E790 storage, all NVMe storage

Hitachi’s first VSP E series storage system, the VSP E990, was introduced last April and now they are adding two smaller solutions, the E590 and E790 storage systems. Both come in a 2U rack mount configuration and support from 6TB to 361TB of raw internal SSD capacity with up to 24 ,1.9, 3.8, 7.6, and 15.0TB NVMe SSDs and include up to 24 FC (32&16Gbps) and 12 iSCSI (10Gbps) host ports. Some more specifications on the two systems:

VSP E590 storage has 24 CPU Cores, 768GB of DRAM caching, and offers up to 76.8GB/sec FC bandwidth to a host or 22GB/Sec FC bandwidth, and up to 4.0M IOPS (we assume 4KB reads), and
VSP E790 storage has 64 CPU cores, 768GB of DRAM caching, and offers up to 76.8GB/sec FC bandwidth to a host or 32GB/Sec FC bandwidth, and up to 6.8M

Similar performance specs for the E990 with 96 NVMe SSDs, are 204.8GB/sec to a host and 30GB/sec FC bandwidth, and up to 5.8M IOPS. All E-series systems are dual controller systems and all are SCM (storage class memory) capable systems.

The VSP E Series software supports:
VSP Adaptive Data reduction which offers up to a 7:1 total efficiency guarantee as well as, a sight unseen, 4:1 effective capacity guarantee. VSP Adaptive Data Reduction uses AI to select the best time to perform data reduction for any IO that will minimize performance impact and maximize data reduction.
VSP 100% Data Availability Guarantee which ensures that VSP storage systems support always on, data availability. Out of the box, VSP E series systems support 6-9s resiliency or 99.9999% up time.
VSP Storage Virtualization which provides a single management point for external and internal storage as well as providing non-disruptive data migration between external and internal storage.

As the VSP E series are midrange systems, they don’t support mainframe attach. There are many other features of the VSP E Series storage , including multiple BC/DR options, data snapshots and extensive VMware support.

HNAS 5200 & 5300 storage systems

Readers of our newsletters will recall that the previous generation HNAS systems, the HNAS 4060, 4080, and 4100 all scaled linearly in performance by node count, and that they supported up to 2 nodes/cluster, 4 nodes/cluster and 8 nodes/cluster, respectively. Also, the high performance of the HNAS systems were in part due to their use of FPGAs to perform compute intensive data services.

The HNAS 5200 and 5300 support a maximum of up to 4 nodes/cluster and 8 nodes/cluster, respectively. Per node performance has also improved:

HNAS 5200 supports up to 3.4GB/sec (sequential?) read bandwidth per node and 1.3GB/Sec (sequential?) write bandwidthper node, and up to 87.5KIOPs/sec (4KB blocks?) per node .
HNAS 5300 supports up to 4.8GB/sec read bandwidth per node and 1.8GB/sec write bandwidth per node, and up to 125.0KIOPs/sec per node .


The new midrange all NVMe, VSP E series storage seems a performance winner to me. Although something producing MIOPS and GB/Sec in bandwidth seems a bit more than midrange to me. The E990 could use a refresh as it offers less MIOPS than the E790, but this could be just due to its lessor core count, (which could easily be solveable).

As for the new HNAS systems, it’s a welcome, if not long-awaited update. HNAS dependance on FPGAs to provide better performance have a downside and that is, it to takes too long to update. The challenge is that unstructured data is becoming even more important to the enterprise and as such, is becoming even more competitive. The end result is that HNAS can no longer hold its competitive edge, if Hitachi continue to take this long to update systems.

Hitachi had the same problem with their original VSP series, which had many ASICs and a monolithic controller architecture with a large shared memory cache. But over time, Hitachi has moved away from this mostly proprietary hardware IP and have transitioned VSP storage to more standard, off the shelf hardware. Something similar needs to be done for HNAS.

[This storage announcement dispatch was originally sent out to our newsletter subscribers in December of 2020.  If you would like to receive this information via email please consider signing up for our free monthly newsletter (see subscription request, above right) and we will send our current issue along with download instructions for this and other reports. Dispatches are posted to our website at least a quarter or more after they are sent to our subscribers. ]  

Silverton Consulting, Inc., is a U.S.-based Storage, Strategy & Systems consulting firm offering products and services to the data storage community

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.