beckmarkwith - Fotolia
Violin Memory was one of the first all-flash storage vendors, shipping its first array in 2008. But being early hasn't helped the vendor stay competitive as the all-flash array market gains momentum in 2016.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Violin has struggled to win buyers for its flash arrays since going public in 2013. Violin reported slightly less than $11 million in total revenue in the final quarter of 2015 and its stock price has dropped so low that the New York Stock Exchange is threating to de-list the company.
Kevin DeNuccio, who took over as CEO of the struggling vendor in February 2014, is in charge of turning the company fortunes. In this interview, DeNuccio discusses how Violin misread the all-flash array market and lays out his plan to return the company to profitability around its Flash Storage Platform (FSP) arrays.
Kevin DeNuccioCEO, Violin Memory
Given Violin Memory's struggles, why did you agree to take over as CEO?
Kevin DeNuccio: Prior to joining Violin, I was on the SanDisk board for four years and had watched the flash market develop. I seized the Violin opportunity because flash is bringing major technology disruption. I liken it to when mainframes gave way to mini-computers, and then mini-computers ushered in the era of personal computers. The same thing is happening with flash storage today. You have Western Digital buying SanDisk, Seagate's stock has been way down over the last six months or so, EMC is selling out to Dell … it's a very disruptive marketplace.
We were the inventor of the flash market. Violin Memory flash storage is different from a lot of the players in the marketplace in that we started with a hardware architecture and, over the last couple years, have added a robust software offering. We think it takes a combination of hardware and software to address the needs of the market going forward.
Violin Memory arrays have lost ground to competitors, and your sales have been down year-over-year. Outline your strategy for turning things around in a highly competitive all-flash array market.
DeNuccio: The flash marketplace is entering its third phase. The first phase was about solving bottlenecks created by disk. We were selling flash for $20 or $30 a gigabyte by inserting our appliances in SAN environments. The cost of flash didn't matter to customers, because it solved a bottleneck that couldn't be solved any other way.
The second phase occurred when [EMC,] XtremIO and Pure Storage came out with inline data deduplication. That changed the No. 1 deployment scenario for flash, which [to that point] was virtualization. Violin became completely uncompetitive [because] it didn't have deduplication software [until recently]. We could get 10:1 to 15:1 ratios of efficiency in virtualized environments. But without deduplication and compression software, it effectively meant our competitors were one-tenth to one-fifteenth our cost. That's what got the company in trouble.
We needed to re-establish the company for the next phase, which is flash moving into mainstream primary storage. Tier zero, tier one and tier two workloads are being collapsed within the enterprise to an all-flash infrastructure.
So, how will you make up the ground you lost to EMC, Pure and others?
DeNuccio: Our architecture allows us to do multiple things. Most vendors have to write to SSDs. Our Violin Intelligent Memory Module (VIMM) flash storage lets you write in the background and start reads from the system without going into garbage collection. Secondly, we break up the I/O and write to the entire array of flash in parallel. Data always resides in the VIMM. We dedupe it, compress it and write it one time so it doesn't need to be moved.
We have RAID groups of four VIMMs and a fifth VIMM as a spare. The system intelligently can see if a VIMM is struggling and automatically copies that data to another VIMM. That intelligence of being able to manipulate our raw flash is very different than an SSD, which is basically a closed package.
You added dedupe and other storage software features with the Flash Storage Platform in 2015. You're still waiting for a sales jump, but how is that array platform being received in the all-flash array market?
DeNuccio: When we introduced the FSP, we set out to add the functionality we needed to compete in serving high-end enterprises with mission-critical apps. We've been in the market now over a year and customers are getting comfortable with it, starting big rollouts to replace disk in their primary storage environment.
By definition, these are large enterprises that are very cautious and move slowly. It takes them about nine months to do a proof of concept. Sales of FSP went from $1 million in its first year to $17 million in 2015. We're projecting year-over-year revenue growth in 2016 between 25% and 35%. The take-up rate is still not getting the company back to its (former) size, but it is starting to drive the growth rate this year. The best example is we've been able to capture about 60 very large enterprises around the world, including 10 global Fortune 500 and about 10 Fortune 100 customers.
Which types of companies are using Violin arrays, and why?
DeNuccio: It's enterprises with big legacy applications that can really benefit from deduplication and compression. They are mainstreaming these applications, which are the heartbeat of their businesses, on flash to get the performance advantages. We have telecom and cable companies, hospitals, insurance companies and retailers all running core database applications for things like billing, claims [and] inventory management. One large retail customer used to be able to only analyze inventory on a weekly basis. Since they started running on FSP, they said they are able to analyze inventory nightly to make distribution and pricing decisions every day of the week.
One of our biggest Fortune 100 customers told us it expects to save somewhere between $15 million and $20 million per year by collapsing multiple tiers of storage on a single platform that can run at very high speeds for traditional database applications, but also with deduplication and compression for virtualized next-gen applications.
Where are most of these customers based?
DeNuccio: We started to win (many) big decisions in Turkey, Russia and Korea by getting CIOs in those countries to talk to each other. That's what we're trying to get going in the U.S. We know we have several Fortune 100 companies that want to go big with us, and we've got them talking to each other.
Acquiring customers takes an investment in sales and marketing, yet Violin recently implemented layoffs. How do you balance the competing pressures of growth while downsizing?
DeNuccio: It's walking a fine line. We do think we can turn the momentum around significantly in the current year. We just finished the first quarter and believe we can turn the corner and start to grow the company again, at the kind of rates it had when it was flourishing. But we also believe it's critical that we get some go-to-market partnerships.
Which kinds of partners does Violin Memory need?
DeNuccio: There are server and network vendors that don't have a storage strategy. At the high end, the key growth is in converged infrastructure. We have unique advantages in a converged rack that are demonstrable and differentiated compared to our peers. It's our belief we will live in a land of significant strategic partnerships over the coming months.
Violin hired an investment bank to consider acquisition offers. Nothing came of that, but is Violin still in acquisition mode?
DeNuccio: We're a public company, so I always say that we're for sale every day. Going through that process is very difficult. Announcing a public process and then not getting to a conclusion where you sell the company certainly has put a heavy weight on the stock. The market is so disruptive right now that Wall Street has trouble predicting who the winners and losers will be. As we went through that process, we got close to several companies that weren't ready to make an acquisition, but they are ready to partner. The end result is we think we'll be able to get at least one strategic partnership out of it to help us scale the company.
Better understand the all-flash array market with this guide
All-flash products are going to be big in 2016
Six of Violin's competitors in the all-flash market