CEO talks Pivot3-NexGen integration, HCI growth

Pivot3 CEO Ron Nash said integrating NexGen Storage's QoS features will increase the number of applications that can run effectively on hyper-converged storage.

This has been a year of growth so far for hyper-converged vendor Pivot3.

The vendor merged with PCI flash vendor NexGen Storage in February. The Pivot3-NexGen deal gave Pivot3 quality of service (QoS) and dynamic provisioning technologies as well as doubled its engineering staff and expanded its go-to-market team.

Pivot3 CEO Ron Nash said his company also increased revenue by more than 50% quarter-over-quarter. That's significant because storage vendors usually experience revenue declines in the first quarter of the year.

Nash said the revenue spike is due to the rising interest in hyper-convergence, publicity from the Pivot3-NexGen merger, and customers who bought its technology for video surveillance and are upgrading to run more applications on the hyper-converged systems.

We spoke to Nash about these developments, and new products coming out of the Pivot3-NexGen combination.

"We had a nice fourth quarter but an even better first quarter," Nash said. "We also doubled the size of the development team. If you have twice as many engineers, that's a competitive weapon. We've started to distinguish ourselves as one of the leaders in the hyper-converged market."

Ron NashRon Nash

How is the Pivot3-NexGen integration working out?

Ron Nash: The big thing is we have the capability to take their technology and our technology to come out with joint products that have the best of both. Our first product will be announced this quarter, and we have another one coming out in the late third quarter or fourth quarter. NexGen had a quality of service and dynamic provisioning capability. On an application-by-application basis you could classify applications as highly critical, regular criticality or low priority, and you put them in categories. That ensures that you always get the response time you need on the most critical applications.

We wanted to put that layer over our hyper-converged products, and also extend it to cloud. So you can have cloud storage, some hyper-converged, and regular storage, and you can have one layer that assures you of a level of service by application across all of that. That's one of the products we got from NexGen and we are putting that together right now with our hyper-convergence. I think it's going to be a killer product for somebody going to the software-defined data center.

So much of hyper-converged has been sold as a platform for a single application, but our customers are using us for multiple applications. Having that quality of service and dynamic provisioning layer will be really important. If you're going to run a giant data center with 200 applications, you have to treat them differently. Being the first hyper-converged company that has something like that is terrific for us. It distinguishes us. We have a fundamentally different technology that gives us an advantage.

You say you have established yourselves as a leader in hyper-converged. What are you basing that on?

Nash: We have 2,000 customers installed and we're in 53 countries. So, it's partly based on revenue momentum. The other thing happening -- and this is a little more subtle -- but when we first started Pivot3, one of our first use cases was storing video surveillance data. We store large data sets well. We also can really ingest a lot of input fast. A lot of storage systems can't ingest for video storage without dropping frames and losing parts of the data. We had a ton of customers buy us for video surveillance, and those people who bought us just for video surveillance are now coming back for more use cases and more applications on our platform. That is also driving revenue up.

Say hyper-converged now and people know it. But talk about different kinds of hyper-convergence and what it really does, [customers are] still in early adopter phase.
Ron NashCEO, Pivot3

What are the common use cases they're deploying Pivot3 for now?

Nash: Virtual desktop infrastructure (VDI) is one. But a lot of people are using us for basic databases, and backup and recovery. We're still selling new video surveillance systems, but we have customers with eight, 10, 12 applications running on our platforms. A lot of databases, ERP [applications] and stuff like that.

Who do you usually compete against in deals?

Nash: About 75% of the time we only run into incumbent vendors selling old products. So, it will be EMC or NetApp storage, or [Hewlett Packard Enterprise] and Dell servers, but we're the only hyper-converged product at the table. In the other 25% where we have another hyper-converged competitor, it's probably Nutanix 80% of the time. The other 20% is all over the map with no dominant competitor. Nutanix is the only one we see regularly.

How do you fare against Nutanix?

Nash: It depends upon how closely [the buyer] looks at the technology. If they look at our technology very closely -- or if they do a proof of concept and put us and Nutanix head-to-head doing a VDI test or a database test -- our win rate is more than 80% percent. If they just look at the marketing hype and literature and say, 'I see Nutanix everywhere and they're advertising and every analyst is talking about them,' then we have a low win rate. So, we have to continue to articulate that our architecture and technology is fundamentally different than that of Nutanix.

What is the most significant architectural difference between you and Nutanix?

Nash: The big differentiator is we're more sophisticated on the storage side. We started out as software-defined storage, and doing software-defined storage is very difficult. We had a 35-man development team that had delivered seven generations of product for Compaq Computer. It took that team five years to get the software-defined storage side working, then we spent one year after that to add the processing side to make it hyper-converged. Nutanix started, and the next year they had a product. They just did the processing side. They haven't done the storage side.

We use erasure coding -- a super set of RAID. Instead of replicating it, we put an algebra formula against the data and only hold results of that formula. That allows us to re-create data if we lose it. Ninety-four percent of our physical storage can be used on unique data.

When you're more efficient on the storage side, you don't chew up CPU cycles. Nutanix has a file system; it works most efficiently when the data and application are on the same machine. We have a SAN; we work no matter which machine the data is on.

Are you close to profitability yet?

Nash: Well, we're a whole lot closer than someone like Nutanix or Pure Storage. They're on a different business plan than we are -- raise massive amounts of money, hurl it into sales and marketing, and race to an IPO before you run out of money. And that can work, but I just wouldn't sleep at night running that way.

We were headed toward profitability, but then we made the decision to acquire NexGen. We just took some additional money from investors for the Pivot3-NexGen merger, and that took us farther from profitability. We are not hemorrhaging money anywhere near the magnitude those other companies are losing. We're losing some, but it's well under control. Our investors understand where we are and are ready to back us to get to profitability.

When you talk to prospective customers, do they already know about hyper-converged or do you still have to explain the concept?

Nash: About 80% of customers know hyper-converged now. Thank you, Nutanix, for spending millions of dollars in marketing. It's great that they did. They educated the market and I'm thankful for it. Say hyper-converged now and people know it. But talk about different kinds of hyper-convergence and what it really does, they're still in early adopter phase. They aren't knowledgeable about the differences in one product from another. The market still has to mature about that.

Next Steps

New platforms continue to populate the hyper-converged market

Hyper-converged and converged infrastructure: What's the difference?

How to implement hyper-convergence without going all-in

Dig Deeper on Hyper-Converged Vendors and Products