Qumulo CEO Bill Richter: Single-Sourced Storage Hardware ‘Very Old, Tired Concept’

“We now see large-scale customers saying, ‘Hey, I want to operate more like a public cloud does. I want standard commodity hardware. And I want incredibly smart, extensible software to power my applications and my people and data.’ And that’s a completely different mindset from what the legacy industry has been doing. [We] don’t believe in rigid, single-sourced hardware to find solutions because that’s really a very old, tired concept,” says Qumulo CEO Bill Richter.

Looking To Manage Burgeoning Unstructured Data

Managing unstructured data, or file and object data that is not stored with a pre-determined structure such as text or video, is becoming a major concern for businesses with large data stores. Research firm Gartner in 2023 estimated that large enterprises by 2028 will triple their capacity of unstructured data across public cloud, on-premises and edge infrastructures compared with 2023.

By 2028, about 70 percent of file and object data is expected to be deployed on consolidated data storage platforms, compared with 35 percent in 2023, Gartner also said.

The data storage industry is responding with a variety of new file systems designed specifically for unstructured data. One vendor, Seattle-based Qumulo, late last year unveiled its new platform, Azure Native Qumulo, that allows businesses to access hundreds of petabytes of unstructured data service on a public cloud.

CRN recently spoke with Qumulo CEO Bill Richter to discuss not only the new Azure Native Qumulo and other of the company’s recent technology introductions, but also about the importance of software to the unstructured data business.

“We’ve spent the last 10 years optimizing organizations’ ability to create, manage and curate their unstructured data,” Richter told CRN. “What’s more notable is what I didn’t say Qumulo was, which is that we build a storage device or we have a rack of stuff that does interesting things. For us, it’s all about software, and that software runs anywhere. Because people know data and applications span public clouds, data centers and the edge.”

While Qumulo works with Microsoft Azure and Google Cloud Platform, the focus on Azure with its latest release was no accident, Richter said.

“What’s different about this is the joint development with Azure and the fact that it’s a fully managed service,” he said. “So customers don’t run Qumulo software on Azure, they consume Qumulo data services on Azure. And that was something really important to Azure.”

Here is more of CRN’s discussion with Richter.

How do you describe Qumulo?

We think about three pillars: people, data and applications. And the problem that we solve is at the intersection of those three pillars. And we’re passionate about it. And when those three things—people, applications and data—work together, there’s enormous breakthroughs for organizations. And that’s at the highest level what we really obsess [about] here at Qumulo. Underneath that, of course, is the notion of unstructured data and its impact on the world and the enterprise. And so from a technology perspective, we’ve spent the last 10 years optimizing organizations’ ability to create, manage and curate their unstructured data. What’s more notable is what I didn’t say Qumulo was, which is that we build a storage device or we have a rack of stuff that does interesting things. For us, it’s all about software, and that software runs anywhere. Because people know data and applications span public clouds, data centers and the edge. And our mission is to help those organizations across all those environments.

But what does Qumulo do with the people, data and applications?

The product is a platform that allows organizations to create, store, protect and curate their data. And what’s unique about us is we’ve observed the real problems that customers are facing in storage, as well as old battles about scale-up versus scale-out architectures and different ways to build storage systems. And we’ve created what we believe is a new category of storage we call ‘Scale Anywhere.’ We believe the concept of storage really needs to be scaled anywhere. And we defined that as being able to power those applications, or allow people to power those applications, that produce enormous amounts of unstructured data in their corporate data centers, on standard hardware, at their edge deployments and, importantly, in public clouds.

In November, we introduced an enormous revolutionary step forward to allow customers to do that. Part of the news is that Qumulo together with Microsoft built the industry’s leading unstructured data service on any public cloud, called Azure Native Qumulo. Customers can scale to tens and hundreds of petabytes on Azure as a native service. It’s completely elastic. It scales up and down. We’ve separated data persistence from compute, so customers pay for those separately only if and as they consume them. And from an economics standpoint, it’s 75 percent less expensive than any other scale-out service on any public cloud. And so it redefines economics for file data or unstructured data in the public cloud.

We also announced a global namespace that will allow customers to have a common data plane between their Qumulo environments, whether they’re at the edge or the core or in the public cloud. A global namespace is really a unified data plane for true hybrid applications.

Anything else?

Also new was a capability that we call ‘Run Anywhere,’ which allows customers to run Qumulo on any standard x86 hardware. We already today offer a very robust hardware compatibility list. And with this next version of Qumulo, that widens out significantly, allowing customers to effectively bring their own hardware. This is super important, especially as supply chains are moving rapidly and customers’ preferences change around not only their hardware vendors but also the capability of the hardware. We support everything from the most screaming-fast NVMe all-flash environments to super economical, incredibly dense platforms to meet customers’ needs of where they are on the price/performance continuum.

We also unveiled Nexus, which is a single pane of glass for management. So if you have a unified data plane across environments, it’s natural that customers will want a unified management plane. And that’s what Nexus is. We have so many customers that are running six or eight different Qumulo environments around the world, both on-prem and in the public cloud, that they can run that from a single pane of glass with Nexus.

Finally, we introduced Qumulo One, which is a commercial model that allows customers to pay for Qumulo on the data that’s consumed, not up front for provisioned capacity or raw capacity.

What would you say makes the platform ‘industry-leading?’

Well, no one can beat our capability, our performance and our economics. And that will become apparent at the moment customers log in to Azure and they see the performance and prices that they will benefit from. I mean, it’s far and away differentiated from anything else that’s out there. And we’ve been able to build that in concert with Microsoft. So we get the benefit of engineering directly with the Azure teams to take full advantage of their best capabilities combined with our best capabilities. And I call that ‘industry-leading.’ There’s nobody else out there that’s able to offer that to customers today.

What part of Qumulo’s business goes through indirect channels versus direct?

Virtually 100 percent of our business involves a partner.

We ask because, if a customer works directly with Microsoft on Azure, do Qumulo partners still get exclusive access to that to offer to their customers?

Qumulo partners absolutely can get access to the Azure service. Microsoft has a very robust partner community that has full access to this service. And the same goes, by the way, for AWS. And then of course, in the data center and at the edge, we rely heavily on our partner relationships to add value together for customers.

Since this is coming through Microsoft, could somebody use a credit card to get it directly for themselves?

They can. But here’s what's really happening. At virtually all organizations now, with a few exceptions, if you ask the CIO if they use a data center or the cloud, they will immediately respond and say, ‘We are hybrid. I use public clouds for these purposes. I use data centers for these purposes. And I have edge environments.’

And what those customers are looking for is a unified platform across all those environments. And what they’re asking their channel partners to do is help them architect those solutions, help them find the right tools to be able to accomplish the real goals of hybrid IT. And if you think about the real goals of hybrid IT, it’s not just to say, ‘Well, I have some cloud and I have some data center.’ What they’re really trying to do is find the right environment, the right tool, for the right job in service of their organization. And resellers can offer enormous amounts of value here because they’re not doing it for the first time. They’ve been able to see patterns emerge and then bring the expertise they’ve developed to customers. And I think customers are really looking for that value now. And so for us, that’s a great fit for the partner community.

One really important thing about Qumulo, especially for our partners, is when the customer says, ‘Oh, those applications or that data set—I want to put [them] in the public cloud,’ our partners no longer have to say they don’t have a role in that. Now they get to say, ‘Because I’m reselling Qumulo, I’m extremely relevant in that conversation.’ That’s important for the customer, obviously, but it’s also important for resellers to be able to innovate on behalf of customers.

Does Qumulo expect to do the same with other clouds including Amazon Web Services or Google Cloud Platform?

Qumulo has for many years offered products on AWS and GCP. And we will continue to because there’s obviously robust demand there. What’s different about this is the joint development with Azure and the fact that it’s a fully managed service. So customers don’t run Qumulo software on Azure, they consume Qumulo data services on Azure. And that was something really important to Azure. They have shared with us that their customers want to be in a full consumption services model. And there was recognition in Azure that they had no scale-out file offering in their catalog, and based on a lot of customer demand, Azure picked Qumulo to develop that service with them, and we’re really proud of that.

You talked about this being available on bring your own hardware. How important is hardware to this new technology? Or is it something that can be done without hardware?

Data, especially in the data center and at the edge, has to run on hardware. And the hardware is incredibly important in that equation. There’s no question about it. What’s different about Qumulo versus the other vendors is if I’m buying from the legacy folks I just have one choice. It’s whatever their supply chain happened to produce. It’s expensive. It’s rigid. It represents a single point of failure for them. And what we saw during the supply chain crisis of the last couple years, you’re only as good as your vendor’s supply chain. We now see large-scale customers saying, ‘Hey, I want to operate more like a public cloud does. I want standard commodity hardware. And I want incredibly smart, extensible software to power my applications and my people and data.’ And that’s a completely different mindset from what the legacy industry has been doing. So we are well aware of the importance of hardware. In fact, we obsess over our optimization around hardware. We just don’t believe in rigid, single-sourced hardware to find solutions because that’s really a very old, tired concept.

Could you see making this scale-out offering available through AWS or GCP natively in the future?

Absolutely. There’s no question we will continue to push the bounds of our vision, which is ‘Scale Anywhere.’ And in that category, there’s no question that we’re a leader. We have 1,000 customers that run Qumulo, either in their data center or their edge environments, or in one of the three public clouds. We’re making huge innovation steps with this new technology on Azure. And customers continue to push us in that direction because they want ‘Scale Anywhere.’ So we will continue to innovate our Azure service. And in the fullness of time, I expect to have similar services on the other public clouds as well.

What are some of your strategic priorities for 2024?

Everything for us starts with the customer. And we have gotten fantastic feedback from customers that they love the Qumulo model. They love the flexibility of it. They’re very excited about the concept of Qumulo One because they are now asking themselves openly, ‘Why do I pay for things before I use them?’ And so you should expect, from a priority st’andpoint, for us to continue to lead the industry around the concept of ‘Scale Anywhere and push the bounds of that. What we find is that when we put ourselves in our customers’ shoes and walk in them for a day, we end up interested in solving completely different problems than we may have expected. And that’s led us to some fantastic breakthroughs. That’s what led us to this notion of Scale Anywhere and being able to run Qumulo in all environments. It’s what led us to the concept of Qumulo One and allowing customers to pay for just what they use. It led us to the concept of offering a global namespace or a unified data and management plane. And so we will continue to press on those bounds in service of our customers. What I find as CEO is when we solve customers’ problems better than anybody else does, sales and customer adoption naturally follows. That served us well since we started the company. And I don’t expect to change that anytime soon.