Transcript
Today we're going to talk about the core difference between Verge I.O. and really just about any other VMware alternative solution on the market. And that difference is our integration, right? If you look at the normal stack that you see in solutions, we'll pick on VMware a little bit. But, you know, at the center you have ESXi, below that you've got vSAN, above that you've got vCenter, and above that, in some cases, you'll have NSXT, which is their software-defined networking capability. And so all of this is now squeezed into a single box that, you know, we would call hyperconverged architecture, and as you scale the environment, add more servers, you do so, and each one of these will get a portion of the stack. And this really creates the problem, right? First of all, it is a stack, that each of these layers essentially run as a VM, or multiple VMs in some cases, to the ESXi hypervisor. And so what happens is you have a lot of coordination that has to happen back and forth between these modules, right? And that causes, you know, impacts performance and things like that. You also have some inner node issues, as each of these now has to have a separate lane of communication between them in order to communicate. Now, what VergeIO does that's fundamentally different is we take a different approach. And what we do is, well, we started with our storage software, our vSAN, that we call VergeFS. It's written from scratch by us, designed very specifically to work in a virtualized environment, and also to be integrated into, not be a VM of, the hypervisor. And so to do that, we used KVM and QEMU to essentially create VergeHV, which is our hypervisor. But the code is actually injected, if you will, inside of it. This gives us a couple advantages, number one, an obvious efficiency gain, things like that, and also a really good understanding of KVM and QEMU at a code level, which most companies don't actually have. The other thing we did is we also integrated in Verge Fabric, which is our software-defined networking capability, so layer two, layer three sorts of capabilities, so that you can do firewall routing, VPN, DNS, BGP, et cetera, right? But like I said, this is all integrated into a single piece of code. So when you install our product, you're going to install VergeOS, and you're not going to have to install a bunch of VMs in order to get that environment operational. There's three big advantages to our doing this. Number one, we can run on virtually any physical hardware in the environment, any server hardware, let's say, built within the last four or five years, generally works just fine. We have a minimum requirements list for you to double check. And that gives us a lot of portability, so you don't have to replace hardware right away, things like that. We can support a variety of different CPUs, multiple generations, and brands like Intel and AMD. We can also even support GPUs, so for more advanced graphics or AI, ML type of work, we can support and virtualize those also. And then using our vSAN technology, we like to have the storage local, and then we aggregate that storage across multiple nodes. So that's the first thing, very portable, flexible in the hardware it can use. The second thing is ease of learning. What I hear repeatedly from customers is that they're able to learn this very, very quickly. Probably 90% of our customers used to use VMware at some point. Now they can switch to us, and in a day or two, they pick it up very, very quickly. More importantly, operationally, what I hear is they're able to get more done in their day. That task list, they're able to knock it out instead of have it just keep growing on them. So this is operationally more efficient. The third thing is performance. We tend to see a 20% to 30% improvement in CPU utilization, and about a 25% to 30% improvement in storage IO performance. And so all of those add up to a pretty compelling solution. So if we zoom back out here and look at how we'll configure this environment, basically what we'll do is we'll have multiple nodes, and those nodes will be in a single instance, and it gives us the ability to scale quite easily. And then what we're going to do is we're going to take the storage that's inside of these servers here, essentially virtually aggregate that into a virtual volume or a vSAN, and then all the VMs are going to access that virtual volume. The other thing that's powerful about the solution, let's say you've been running along for a while. Things are looking great. You need to upgrade, but you don't need to upgrade both storage and compute. You just, say for in this example, just need more compute. Well, what we could do is you could buy a node that just has essentially a boot drive. Basically you go with four processors, because we charge by the physical server, not by the number of processors. So there's no penalty in doing that. And then you add that to the existing environment. And that's a perfectly valid configuration under VergeIO. This system here would get its storage resources from that aggregated volume that I had mentioned earlier, and so it would work fine and be totally protected and things like that. So that's really the first key part of VergeIO is this integration. As we go through other instructional videos, there'll also be areas where you'll see this integration pay off. But the big three are the ability to run on a wide variety of hardware, so it's very portable. Number two, quick learning curve, very efficient operations. And then number three, improved performance and performance efficiency, both CPU and storage. Thanks for joining us. I'm George Crump with VergeIO. Have a great day.
