Yesterday, Violin announced that they have acquired Gridiron Systems, a vendor of cache acceleration hardware and software. At the outset it may seem odd that an all-flash vendor would want to acquire a company offering acceleration for traditional arrays, however this could be a very smart deal indeed in a number of ways.
Software is King
Although hardware is important (IT has to run on something), most differentiation comes in the software. Violin currently has a different architecture using VIMMs (Violin Intelligent Memory Modules), but they still ultimately use NAND flash. There’s also some DRAM in there used for storing metadata, but the persistent technology is the same. Differentiation can come from software, a message surely not lost on CEO Don Basile, who was formerly the CEO of Fusion-IO, a PCIe SSD card memory vendor. Fusion-IO sells solid hardware, but the effectiveness of software solutions in that space will be the reason that the company succeeds or fails. By acquiring Gridiron, Basile could be following his Fusion-IO strategy of clever software to make use of the hardware; covering all his bases, to use a baseball analogy. But how could he use it?
Flash For Everything
I can see two possible ways that the Gridiron technology could be used by Violin.
- Sell it standalone. Continue to sell the Gridiron OneAppliance TurboCharger as it is today. There is stil a huge market of HDD-based arrays out there that still sit on the balance sheet and could benefit from this technology. In addition, for those organisations who are skeptical about all-flash arrays at this stage, the Gridiron solution offers an entry point into the customer at low risk; the TurboCharger is only accelerating read requests and so could be removed with little impact. What it does do, however is demonstrate to the customer that flash could be useful, and so lead on to an all-flash sale.
- Use the IP within Violin technology. The all-flash market is getting more competitive by the day. Hitachi are making dedicated flash hardware, EMC are entering the market (eventually) with XtremeIO and there’s a wide range of other platforms to choose from – Whiptail, Kaminario, Pure Storage, Nimbus to name but a few. Startups have to keep ahead of the competition and the Gridiron IP could be integrated into Violin’s existing product line. Remember that all-flash arrays are great at write IOPS, so being able to migrate more read workload to a dedicated tier or section of the appliance could allow Violin to squeeze more performance out of their existing boxes and keep that competitive advantage.
Of course Violin could choose either course, or another one entirely, only time will tell.
The Architect’s View
Violin are determined to own the all-flash marketplace and are aggressively pushing their technology forward. In some respects they remind me of EMC in the old days. Adding to their already considerable IP can only help propel them forward and keep that differentiation any startup needs against the big storage vendors. With an IPO for the company imminent, it’s certainly an interesting time indeed.
- Violin Memory Said to File IPO with $2 Billion Valuation (Bloomberg, 17 October 2012)
- Violin Memory Buys Flash Memory Rival GridIron Systems (Forbes, 18 January 2013)
- Violin Memory grabs Fusion-io’s former chief (The Register, 15 September 2009)
- Two New Podcast Appearances – Speaking in Tech and Violin - 27 April 2012
- The Evolution of Solid State Arrays - 11 November 2012
- XtremIO (aka Project X) – Where’s the Innovation? (Updated) - 10 December 2012
- Violin’s $200 Million Acquisition? Don’t Buy It – 22 January 2012
Comments are always welcome; please indicate if you work for a vendor as it’s only fair. If you have any related links of interest, please feel free to add them as a comment for consideration.
Subscribe to the newsletter! – simply follow this link and enter your basic details (email addresses not shared with any other site).
- Netapp: The Inflexibility of Flexvols (9,967)
- Windows Server 2012 (Windows Server “8″) – Storage Spaces (9,404)
- Enterprise Computing: Why Thin Provisioning Is Not The Holy Grail for Utilisation (7,859)
- Comparing iSCSI Targets – Microsoft, StarWind, iSCSI Cake and Kernsafe – Part I (5,831)
- Review: Compellent Storage Center – Part II (5,475)
- Data ONTAP 8.0 – Part III (5,062)
- Why Does Microsoft Hyper-V Not Support NFS? (4,892)
- Back to Blogging (4,451)
- How To: Enable iSNS Server in Windows 2008 (4,348)
- Windows Server 2012 (Windows Server “8″) – Virtual Fibre Channel (4,171)
- Windows Server 2012 (Windows Server “8″) – Virtual Fibre Channel (12)
- ViPR – Frankenstorage Revisited (10)
- How To: Enable iSNS Server in Windows 2008 (9)
- Review: Compellent Storage Center – Part II (8)
- Why Does Microsoft Hyper-V Not Support NFS? (6)
- HP Storage Bets on 3PAR (6)
- Comparing iSCSI Targets – Microsoft, StarWind, iSCSI Cake and Kernsafe – Part I (5)
- Gartner ECB (External Controller-Based Disk) – 4Q2012 (5)
- Choosing Between Monolithic and Modular Architectures – Part I (5)
- Review: Compellent Storage Center – Part I (5)