Qumulo cloud approach gang-tackles file-to-object conversion

NAS vendor Qumulo is moving beyond its startup status. The eight-year-old company recently closed a $125 million funding round that brings its total haul from private investors to $350 million. Qumulo CEO Bill Richter said his company took the additional financing to accelerate product development, particularly to support a surge in unstructured data. Qumulo cloud customers span government, healthcare, high-performance computing and media and entertainment, along with other industries that rely heavily on streaming data.

"We last raised capital [in 2018] and still have about 50% of it still in the bank. We chose to raise this additional round because our market opportunity is so big," Richter said.

Qumulo's founders helped create the Isilon scale-out technology that is now part of Dell EMC. Richter, also an Isilon alum, became Qumulo CEO four years ago. While the Seattle-based company initially broke into storage selling branded appliances, Richter said Qumulo's future growth depends on furthering its software-defined storage approach to data management in the cloud. The Qumulo distributed file system is packaged on branded Qumulo appliances, on Hewlett Packard Enterprise Apollo-branded servers, or licensed in AWS and Google Compute Platform.

The software platform recently expanded to include Qumulo Shift, which allows customers to run native file services in AWS and convert files to data objects that access Amazon Simple Storage Service-based services.

We recently spoke with Richter to discuss how growth in unstructured data shaped the Qumulo cloud roadmap.

You have indicated that Qumulo may pursue an initial public offering (IPO). How will the latest funding haul influence your decision to go public?

Bill Richter

Bill Richter: Our goal to build a long-term independent, sustainable company. I think that journey implies an IPO one day, but we'll go public whenever we're ready. We have been relatively cash-efficient. Based on the way we've built and managed Qumulo, and given the amount of capital we raised, we get to decide when or if we do an IPO. It will not be decided for us.

The storage sector seems more dynamic than ever. What challenges does that pose for Qumulo? What opportunities does it present?

Richter: To understand Qumulo's success, you have to zoom out and look at a couple of mega trends affecting our business. The first is massive and unrelenting digitization of data. The scale of the data sets that we help customers with is many orders of magnitude larger than what is commonly referred to as big data. The massive proliferation of unstructured data is one trend.

A second trend is the advent of cloud computing and software to solve these massive data problems. People used to try to solve this by buying new storage boxes in their data centers. We saw a different opportunity: to tackle the unstructured data challenge by using our file software in the public cloud. That was the early thesis of Qumulo when it was founded eight years ago. Those mega trends have only accelerated and we believe it positions us to build a strong and successful company.

You started out with file storage, then added object support for the cloud. Do you envision expanding to block for unified storage?

Richter: We are highly specialized around file data. We don't provide a block data system. We don't provide a transactional data system. Qumulo is not a backup company. In fact, there's a vibrant conversation about this inside Qumulo. We use this phrase 'focus is your friend.' We focus on file data, which is an overlooked market segment [for cloud workloads]. The incumbents are not innovating in that area, because it happens to be a very difficult problem to solve.

We see customers whose applications are better suited for the public cloud, and those applications all use file data. Another aspect is that many customers aren't sure whether those applications should live on premises or in the cloud. When they buy Qumulo, they buy the optionality to transport an application from Point A to Point B and back again, without having to change the application.

AI data inferencing and cloud data management are fueling demand for faster storage in the cloud. How do these dynamics change the way customers evaluate their storage?

Richter: As a matter of fact, storage is important, right? If you can't store the data somewhere, you can't get started. But storage really is just the starting place. There is an intersection of three things: data, applications and users. Storage is the foundational element.

We are to data storage like a fire truck is to water storage.

We are to data storage like a fire truck is to water storage. Qumulo is all about enabling data-intensive applications to grow at scale. What inferences can you draw from that data? What kind of data services can you offer? How can you move that data around? And how can you get insights from that data? You can do some interesting things by selling a new storage box, but if it were a baseball game, that storage box only gets you to the second inning.

Where is the Qumulo cloud software most taking root? Who are your typical customers?

Richter: The vast majority of our business comes from the lines of business inside an enterprise. When we sell to Fortune or Dow 30 media companies, we are not dealing with the CIO. We have conversations with the creative department. We have conversations with the production department. We have conversations with broadcasters. These are deep discussions about workflows and what customers need Qumulo to do to enable the data-intensive applications those environments need.