Rethinking the VMware Exit: Infrastructure, VDI, and What Comes Next with VergeIO

Episode Transcription

Karen Gondoly, CEO, Leostream

George Crump, CMO, VergeIO

Karen Gondoly:

Welcome, everybody, to another Leostream podcast. I am very excited this time to be joined by our partners at VergeIO. This is George Crump, the CMO.

George, do you want to introduce yourself and give us a little overview of what VergeIO does?

George Crump:

Sure. Thanks for having me, Karen.

We are a company that provides infrastructure software. We converge compute, virtualization, storage, networking, and data protection into a single piece of code that allows people to rethink the way they approach infrastructure.

Right now, of course, with everything going on with Broadcom and VMware, we’re being seen a lot as a VMware exit option. We’re happy to be that, but one of the things we try to get across is that we’re much more than just an exit strategy.

Karen Gondoly:

I’m sure the Broadcom acquisition of VMware made life pretty exciting for you.

How are those conversations going? What are the important things people are asking about when they’re looking to migrate off VMware?

George Crump:

It’s interesting because we’re getting ready to enter our third year of this now.

When it first started, I think there was just shock that Broadcom did exactly what Broadcom has historically done. They raised prices dramatically. I remember in the early days talking to customers that were seeing 400, 500, even 600 percent price increases.

The conversation has matured since then. There were clearly a number of people who switched right away, and then there were a lot of organizations that did a short-term renewal, maybe for a year or two. We’re starting to see a lot of those renewals come to fruition now.

And they’re asking good questions. How do I migrate? What’s my outage window going to be? How am I going to protect the new environment? How am I going to handle disaster recovery? How do things like virtual desktops work in this new environment?

And one we’ll hear occasionally is, “Should I go to the cloud?”

Those are the kinds of questions we’re seeing most often today.

Karen Gondoly:

When I think about VMware, because Leostream is a connection management and remote access platform, I tend to think in terms of VDI and the hypervisor.

But there’s so much more that goes into the infrastructure. So, the conversation is going beyond just vSphere. You’re talking about storage and a number of other pieces of the infrastructure as well, I imagine.

George Crump:

Absolutely.

Because we address all of those different components, the conversation usually starts with, “Can you really do what VMware does? Can you do things like live migrations?”

And the answer is yes.

Then it becomes a broader conversation around how storage is different, how networking is different, and how data protection is different. Hopefully, we have good answers for all of those things.

But yes, it’s definitely a broader conversation.

And then, over the last six months or so, we’ve also seen costs going up in a way that we really haven’t seen before in technology.

Karen Gondoly:

There are two things I take away from what you just said.

First, it sounds like moving to VergeIO creates an opportunity to simplify because you’re bringing all of these pieces of infrastructure together.

Then there’s another opportunity to continue using the hardware investments you already have and save money, particularly when hardware is harder to come by and becoming more expensive.

George Crump:

Absolutely.

Everybody says their product is simple and will simplify things, but I don’t think anybody in IT really believes it until they see it.

That’s usually one of the things we hear very early on from customers. They’re almost surprised by how easy it is.

Because we have that single code base, you’re able to move through these different functions seamlessly. You click and suddenly you’re managing networking, then you’re doing something else. The ability to move through all of that from one place is particularly powerful.

We’ll have customers tell us they’re getting their job done in half the day, or even a quarter of the time it used to take.

We recently had a big customer win with Formica, the countertop company. On the hardware side, they’re now projecting that their current server farm can last until 2034 based on their projected demand and needs.

That changes the way you budget for infrastructure.

Karen Gondoly:

Definitely. That’s amazing.

Let’s switch gears a little bit. You mentioned the cloud. How does cloud factor into this?

Some organizations are still considering it, but what are the big decisions they need to think about when choosing between cloud and on-premises infrastructure?

George Crump:

It’s interesting because, if you look at the roots of our company, we really started out focused on what I would call regional cloud service providers. These are companies that, for some reason, wake up every morning and decide they want to compete with Amazon, Google, and Azure. God bless them.

We were the software that hopefully gave them a fighting chance.

The VMware opportunity came afterward as we started getting the attention of enterprises that have very similar needs and are almost service providers in their own right.

Because of where we started, we have many of the types of capabilities you would expect to see from AWS or Google.

We actually pick up a few customers every quarter that are leaving the public cloud.

It feels like the cloud providers haven’t fully been hit by the increases in hardware costs yet because they had a lot of inventory, but I think they’re getting ready to feel that as well.

What customers notice today are all of those recurring costs that just start to add up.

We run Salesforce.com, for example, and for some reason that ends up on my budget. You look at some of those costs and start thinking, “I feel like we could write our own software for that amount.”

I think a lot of people are starting to have that same thought process.

If you remember the old “cloud first” mantra, I think we’re moving toward more of a “cloud right” approach. Put workloads in the cloud when it makes sense.

Obviously, we have strong opinions on that, but we’re seeing organizations balance their workloads much more frequently than they used to.

Karen Gondoly:

That makes sense. Cloud where it makes sense, rather than cloud just for the sake of cloud, because those costs can become just as expensive, if not more expensive, than the hardware.

I’m also wondering how many of those conversations are focusing on data sovereignty. That’s a big topic we’re hearing about from customers these days.

George Crump:

Absolutely, especially in Europe. We see a lot of conversations around data sovereignty there.

I think AI is going to push that conversation even further.

We use AI with our own code, for example, and I can guarantee you our code is never touching a public cloud. We’ve built our own internal AI stacks.

The challenge with AI is that it’s really good at things that are already known. It’s not necessarily as good at things it doesn’t know. Nothing else knows our code, so there isn’t a body of information for it to learn from.

But what AI is very good at for us is testing. It can run 1,500 tests, help find bugs, do regression testing, and things like that.

I think that translates to businesses that aren’t necessarily software developers as well. Every organization has a set of data that is essentially its crown jewels, and they don’t want that data getting out.

I think AI will drive a lot more of that conversation going forward.

Karen Gondoly:

Right. And AI workloads can be incredibly computationally intensive.

A lot of Leostream customers are also working with HPC and high-performance graphics workloads, so let’s talk about GPU support.

Does VergeIO provide comprehensive GPU support?

George Crump:

Yes. We’re officially listed by NVIDIA as a supported platform. In fact, I had one of the NVIDIA engineers on a webinar say, “This stuff is awesome.” I wasn’t allowed to keep repeating that, unfortunately.

But yes, we certainly support GPUs.

We also support MIG, so you can hard-allocate GPU resources. The product works very well for customers that are running local models.

We can also pool GPU resources across nodes, which allows you to leverage multiple GPUs across the environment.

And with our multi-tenancy capabilities, you can hard-allocate GPUs to specific tenants.

Going back to the service provider example, if you wanted to provide GPU services to customers, whether that’s for VDI or AI, you could designate certain customers as GPU users and allocate those resources specifically to them through our multi-tenancy capabilities.

Karen Gondoly:

So far, we’ve talked about infrastructure in general. For a lot of people, that means servers. For us at Leostream, a big part of the conversation is VDI.

Broadcom acquired VMware, and the end-user computing business was ultimately spun out as Omnissa. Are many of your VMware replacement conversations also turning into conversations about Omnissa?

George Crump:

Not as much. We actually have more conversations about replacing Citrix.

Citrix has obviously had its own changes over the years. But in some ways, we’re almost the reverse of what Leostream does. Our job is to build the infrastructure platform and make it optimal for whatever solution is going to run on top of it.

We want to get to the point where whatever VDI solution makes the most sense for that particular customer can run on our platform.

I wouldn’t necessarily say we view Omnissa as a competitor. The whole situation became a little confusing after the acquisition and the spinout.

But VDI is certainly an area we’re involved in. I’d say about 10% of the projects we’re working on today are VDI-related at some level.

For very small deployments, it’s basically just a few VMs. But when you get into fleet-sized deployments, that’s where a product like Leostream becomes critical.

Karen Gondoly:

And those are the kinds of opportunities we’ve been working on together. Leostream has had a deep integration with VergeIO for a number of years now through the API you provide, and it’s always been great working with your team on those opportunities.

George Crump:

One thing I’ll admit is that we probably don’t bring up VDI as often as we should because, obviously, we don’t sell VDI ourselves.

So what should my team, and specifically our reseller partners, be looking for?

Should they only be looking for organizations that are already running VDI? Or are there situations where a customer isn’t using VDI today but would be a great candidate because of certain characteristics?

What’s the right guidance you would give my team and our resellers?

Karen Gondoly:

When we see people migrating infrastructure, a lot of it ultimately comes down to the data.

People need access to their data, and they need access to the applications that work with that data. Those applications are often accessed through a desktop operating system, which is what Leostream connects users to.

If your resellers are working with customers who are moving data onto the VergeIO platform and those users need applications to work with that data, it often makes sense to move those applications into VMs running on VergeIO as well.

Applications generally perform better when they’re close to the data.

So, when someone moves their data to a new location, they should also think about moving the desktops and applications that need to interact with it.

Then the next question becomes, “How do I give my users access to those resources?”

That’s where Leostream comes in.

George Crump:

I would think the same thing applies to data sovereignty, right?

If you’re trying to maintain sovereignty over the data but you don’t lock down the desktop, you’ve potentially undermined that whole strategy from the beginning.

Karen Gondoly:

Definitely.

You want to keep the data inside the platform instead of transferring it out to wherever the user happens to be, particularly if they’re remote or moving between locations.

George Crump:

And I think the same applies to AI.

You have that same data challenge. We certainly don’t want our code base scattered across 1,800 different laptops. We want it secured and controlled.

I think businesses working with sensitive data are going to approach AI in much the same way.

Karen Gondoly:

Yes. That’s actually becoming a new use case we’re seeing with people hosting AI workspaces.

A user can log into a dedicated workspace to do that work while keeping it segregated and separated from the rest of the environment.

George Crump:

Are you seeing a lot of that?

Sometimes I feel like private AI is one of those things people talk about more than they’re actually doing. Are you seeing businesses doing meaningful work with AI at this point?

Karen Gondoly:

We’re starting to see some of it.

I’ve heard about it particularly in post-production, where organizations are using AI to enhance editing workflows.

But I think it’s still an area where a lot of people are experimenting and learning, like you mentioned earlier, what AI is good at and what it isn’t.

George Crump:

Another big change we’re seeing, beyond VMware, is the increase in server and storage costs.

And I would think that applies to laptops as well.

When all of this started, I used it as justification to buy the biggest, baddest laptop I could get my hands on. It has 48 GB of RAM and a two-terabyte SSD.

I looked at the exact same unit six months later, and it was $1,000 more expensive.

When I look at most of the people on my team, they’re probably going to hate me for saying this, but a lot of them could probably do just fine with something closer to what I’d call a Chromebook.

So, is there an opportunity for Leostream to help organizations keep their existing laptops longer, but host the actual workspace on a high-end server instead and avoid some of those replacement costs?

Karen Gondoly:

Yes, absolutely.

What we see is that organizations can take those existing laptops and extend their lifecycle for as long as possible.

The laptop essentially becomes the client device. The operating system you’re ultimately interacting with is hosted on VergeIO or another infrastructure platform.

You log into the laptop and use a display protocol to connect to the VM that’s actually hosting your workspace.

So yes, that can definitely help organizations get more life out of older endpoint hardware and devices.

George Crump:

Exactly. Most of these devices can last a lot longer than the traditional refresh cycle might lead you to believe.

Karen Gondoly:

Right. Although, of course, the hardware vendors would still like you to buy new hardware.

George Crump:

Yeah, but nobody can afford it anymore.

Karen Gondoly:

It’s a new world.

Karen Gondoly:

Well, this has been a great conversation. If you were going to leave the audience with one last thought about VergeIO, what would it be?

George Crump:

I think the IT community is in a very unique situation right now because we’ve had several major events happen at roughly the same time.

Broadcom purchased one of the largest vendors in the market. And, for the first time that I can remember, the cost and availability of servers have become significant challenges.

This isn’t necessarily like a flood or a hurricane where you just wait and eventually things return to normal.

I think the way to solve these challenges is to rethink the entire infrastructure.

When the cost of technology kept going down, organizations could almost afford to be inefficient. You could even argue that it was sometimes too expensive to focus on efficiency because everything was just going to get cheaper anyway.

I’ll give you an analogy that will probably date me.

In the early days of networking, there was a big battle between Ethernet and Token Ring.

The IBM guys would get up at a whiteboard, draw a circle, and even draw a little choo-choo train going around it to show how perfectly Token Ring could get everything where it needed to go.

I remember walking up to the whiteboard one time, taking a big marker, drawing a thick line, and basically saying, “Yes, Ethernet is inefficient, but it moves data really, really fast.”

Those philosophies don’t necessarily hold anymore.

Today, you have to be efficient. You have to get more life out of your hardware. And data has become incredibly valuable, so you have to make sure it’s secure, locked down, and not going where it shouldn’t.

I think this is the right moment to rethink the entire infrastructure.

Modernization isn’t about starting over from scratch. It’s about leveraging what you already have and putting the right tools in place.

I think our two companies are a good example of that: using software to get more from existing hardware, extend its life, and ultimately simplify the environment.

Karen Gondoly:

I think that’s a perfect summary.

You hit on all of the key points: security, data sovereignty, and doing more with less. That’s definitely the world we live in now.

George, thank you very much for your time. I look forward to continuing to work with your team in the future.

George Crump:

Thank you.

Karen Gondoly:

Thank you.