International Business Machines Corporation (IBM) Earnings Call Transcript & Summary

Hello, and welcome to Enter the Enterprise Edge, where we will dive into the rise of edge environments in today's enterprise. I am Christian Perry, and I am Senior Research Analyst for infrastructure at 451 Research. And joining me on the panel today are Eric Herzog, Chief Marketing Officer and Vice President of Global Storage Channels at IBM; Misbah Mahmoodi, Senior Director of Product Marketing at VMware; Sudhir Srinivasan, Chief Technology Officer and Senior Vice President of Engineering for Storage at Dell Technologies. So before we jump into the panel session, I'll share some insight and data into how we at 451 Research see the enterprise edge developing. First, why the need for an enterprise edge, then I will dive into the ecosystem on a little deeper level, including where the enterprise edge fits across the spectrum of the broader edge ecosystem. Next, we'll take a peek at where edge processing is happening today. And I'll also share some findings that reflect the role of hyperconverged infrastructure in the enterprise edge. Finally, I'll discuss barriers that we see to future expansion, and then I'll bring our panelists back on board, and we'll continue the conversation. And if time permits, we'll address any of your questions. So let's jump in. The edge, in general, has plenty of definitions depending on who you talk to. To us, edge computing generally refers to computing resources that sit outside of the core zone of public cloud and closer to the end user. And the way we define the enterprise edge, in particular, is IT infrastructure physically located within a nearby data center or a micro data center, a campus wiring closet or a facility location, for example, cage off of the production floor in a manufacturing facility. These locations are owned or leased by the enterprise. The infrastructure here looks very similar to the infrastructure that sits in the centralized or nonedge data center. So why the need for enterprise edge? This is a market that has formed due to necessity. Many of the most transformational workloads and related data cannot be served by centralized computing alone. With the enterprise edge, organizations can, for example, perform analytics on enterprise workloads and store the generated data at that location, all without dealing with potential latency problems from uploading data to the cloud and analyzing the data there or even sending it back to a centralized data center. And because you're not sending the data elsewhere, you don't need to worry about high bandwidth costs or security issues related to public cloud. Enterprise edge deployments also help businesses to stay in line with data sovereignty needs. For example, if data is captured in a certain location, there is no concern about breaking laws or governance structures because you need to send the data out to a public cloud provider in a different location. As organizations grow and they become more global, this will become an especially important point that data sovereignty because many businesses will capture data in far-flung locations. And using that data for business advantage at or near the location it's generated will be critical. Finally, another big driver behind the enterprise edge is business continuity. Centralized computing is perfect for many enterprise workload needs, but if a natural disaster impacts that centralized location, and the enterprise has not expanded to the edge as part of its backup and disaster recovery plan, it can face disruptions across the entire operational path from supply chain to partners, to customers, it goes very far. Spending IT to the edge also helps to offset the reliance upon a single location, along with other disaster recovery plans and strategies. And with COVID-19 permanently changing the IT landscape, there's just a general need to look at new ways to approach IT. The edge for years has been seen as the next evolutionary step for IT, if you will. And COVID is pushing IT teams to consider how the new normal is going to impact their business. Edge will inevitably be a part of those considerations moving forward. By 2025, there will be billions of edge locations with many of them at the device edge, which, as you see here, hold sensors, cameras, drones, embedded products, smart phones and similar devices. But the enterprise edge is a major component of the overall edge ecosystem. The enterprise edge consists of micro data centers with edge specific infrastructure as well as MEC, or Multi-access Edge Computing, which moves centralized cloud computing closer to the customer for IOT, for location tracking services, for analytics and other use cases. Enterprise edge locations are also effective places to pull in and aggregate compute and storage workloads from several device edge locations. An example here would be high-definition video cameras that monitor factory line quality control or public safety or control of inventory shrinkage caused by shoplifting. And think here about the impacts of COVID, where future public locations could have heat-detecting cameras to monitor the temperature of people going into buildings, going into airports or going into other locations. All of this will take tremendous amounts of orchestration and management. And increasingly, we're going to need infrastructure that's designed with all of this in mind. This includes platforms that use Kubernetes to drive efficiency, to drive flexibility as well as hyperconverged infrastructure that eases deployment and management in edge locations. We've also been looking at where the money flows in edge computing. The buyers are many, ranging from consumers to enterprises to content owners and the benefactors are also many, as you can see on the left side of this diagram. There are multiple pads that lead back to edge infrastructure, which innately serves as the core building block for enterprise edge build-outs, of course. But for all of this to work, there will need to be substantial strategic updates to traditional business and technology plans that were originally created to accommodate far less dynamic IT environments, more traditional IT environments. This starts in part with simply building out at the edge. So let's take a look at what's forming the enterprise edge. Our data shows that enterprises with IoT initiatives currently have an average of 3,900 devices deployed to support current IoT use cases. And this is going to increase by an average of 65% in just 2 years. Much of what we currently see in terms of IoT endpoints is data center facilities, IT equipment, camera, surveillance equipment and buildings and other structures. But aside from cameras, we also see plenty of other use case specific devices, including field equipment, retail devices, environmental sensors, robots and more. So there's plenty populating this ecosystem. So when you consider the wealth of endpoints that are out there today, you have to figure there's going to be processing happening outside of core data centers, right? Well, that's exactly what's happening. We found that among organizations to perform additional analysis on data beyond that first pass, just 50% of initial data processing happens at the core, whether that's -- whether that core is public cloud, an enterprise data center or a third-party data center. But all of the other initial data processing is happening outside of the core, either at the near edge or at the edge itself. And after that initial processing, 71% of organizations conduct additional analysis on the data. And still, only 56% of that is happening in core locations. The bottom line here is that there is plenty of activity at the edge, and it's only going to grow. In fact, if we look at the last 3 years, we see that edge processing continues to rise in multiple categories, including IoT data processing at the network edge, edge processing on nearby infrastructure versus on a device and even advanced data analysis, including AI at the edge. What's actually gone down in that time is the transport of raw IoT data to a -- back to a centralized location. So it's fair to us that organizations are finding value by processing at the edge rather than sending things back to their core data center. And here's a look at where customers are telling us their spending at the edge. They're buying edge servers. They're buying IoT gateways, software and other technologies to support their IoT initiatives. Note the large bump between 2018 and 2019 in localized micro data centers. So plenty of growth there. I'll also add that the lines are blurring between the various infrastructure devices themselves at the edge. For example, we found that vendors are essentially loading their gateways with more compute, more storage. And in essence, those gateways become much more like servers than just pure gateways. So there's a lot of blurring and overlap going on in the infrastructure market at edge. We expect to see this trend continue, especially as customers look for ways to extract the most resources possible from the infrastructure that sits at the enterprise edge. Remember, there is not a lot of space at the edge. So whatever can be packed into the infrastructure at the edge and still run efficiently, still run cost effectively will be most beneficial and ultimately most attractive for those build-outs. So as part of our voice of the enterprise customer research series, we track the evolution of hyperconverged infrastructure, which continues to be a disruptive force in the IT arena. One thing we noticed in recent years is that HCI deployments are beginning to expand beyond the walls of centralized data centers. Many of the primary reasons HCI has become popular in the first place are the same reasons why we see HCI as an effective fit at the edge, in many cases, or at least in some cases, but we think it's going to build to many cases. You'll notice here that 14% of customers expect to deploy HCI at the edge in 3 years, up from 9% currently. And we'll do this because of HCI's ease of management, ease of scaling, speed to deploy, cost. One of the inherent challenges with the enterprise edge is the lack of IT staff in those remote locations. But once deployed, HCI rarely needs physical attention. Most IT teams can just set it and forget it or deploy it and forget it. We'll continue to track the progress of HCI at the edge, but we do expect this growth to continue. So as I've discussed, the enterprise edge carries plenty of potential to be transformative for businesses. Many use cases are popping up, but results are not guaranteed. First, unless there is a clear use case, for example, let's say, capturing and processing data from devices throughout a supply chain to optimize operations for an organization, for an enterprise. Some organization -- if that case isn't very apparent and clear, some organizations can struggle to develop a clear business case for investing in edge expansion and often, it does come down to money and budget. And speaking of business cases, IT budgets continue to be tight. And every last dollar must be justified and show a return on business value. There are also concerns that data in edge locations might be at risk of breaches. There's also a lack of internal expertise or rather, and more importantly, a perceived lack of internal expertise because, again, this is infrastructure that closely resembles edge -- enterprise edge infrastructure closely resembles what's already in centralized data centers. But there's sometimes a perception that the edge requires special infrastructure skills and knowledge, and that's not necessarily the case. So there needs to be more awareness and education built around what's actually available for the edge and how it's deployed. And finally, there are also concerns that edge infrastructure won't match with existing infrastructure and that deployments can be overly complex. And in many cases, that's true, especially for larger deployments. So what's the bottom line here? The enterprise edge is expanding and it's really -- it's about more than just gathering the first round of insights from data captured in those edge locations. When an enterprise expands into the edge, it tends to use locations to receive -- these edge locations to receive and aggregate data from multiple locations. So in a sense, these become fully functional data centers or, in this case, micro data centers that deliver tangible business value. There's true value here as long as there's a use case and everything is put in place to support that use case. I mentioned in the barriers section that simplicity will be key. Even if there is a clear use case, the path to achieving that use case can be murky, if the infrastructure deployment is difficult. And finally, awareness will improve around the enterprise edge. And as it does and as we see deployments continue, we'll start seeing uptake of associated technologies such as Kubernetes and others into the ecosystem to sort of broaden the functionality and essentially bring those remote sites more efficiently into the realm of core IT. Okay. So what's next with the enterprise edge? To find out, I'd like to open up the discussion to our panelists. So panelists, please open up your camera and your mic, and let's get started.

Okay. So let's start with Sudhir. How do you see the enterprise edge shaping up? What has sort of captured your interest in this IT segment?

Yes. Christian, that was -- thank you. Thanks for having me, by the way. And it was a wonderful overview. I think you really laid out very well. We see the enterprise edge very much in line with what you've got. The taxonomy makes tremendous sense. I noticed you didn't have any sort of sizing numbers, and everybody's got their own sizing numbers. There's a lot of numbers flowing around. But depending on who you ask and how you count, this is where the action is in the next decade, is hell of a lot of compute that's going to happen at the edge. And in storage as well, although storage has quite a different flavor to it at the edge than it does at the -- what we've been used with the core data centers in cloud. But we basically see this as this is where all the growth is going to happen or most of the growth is going to happen in both compute and storage. The devices, for sure, make the hardware infrastructure, et cetera, but I think the thing that unlocks it all is the software. And that's where we see the tremendous amount of opportunity is in the software stacks that run at the edge, the management software that makes it all easy to operate, like you said. The one big difference being when you're in a data center or 2 data centers or even 5, it's relatively easy to manage. Now we're talking of thousands of points of presence and managing all of that in a secured way without a perimeter is a whole new ballgame. So I think that and then the software around analytics itself is just -- that's where all the action is going to be.

Yes. I think if you take any IT ecosystem, the larger it gets the more complex it gets and the more difficult it gets to manage. And here, I think we have the extreme example with enterprise edge in terms of growth and in terms of scaling. So yes, I can definitely see how a lot of those things are in play. Misbah, let's go over to you. What's your general take on the enterprise edge and the state of the market?

Yes. So thanks, Christian, for having me over here. One of the things that you highlighted and I thought was it's really important as enterprises look at edge applications or edge, in general, is what are the specific applications and use cases that they want to be able to deploy because that will actually dictate where you're actually going to be locating various different compute storage networking resources. And so at VMware, we're thinking about it from multiple different use cases, whether the enterprise is looking at manufacturing use cases, retail, health care, even broader telco-type use cases. And what we have found is that different use cases, there are specific requirements that will require you to locate compute storage in different areas of your network. And so let's just take, for example, manufacturing. You would want to have your analytics, your reporting, some of the controls in functionality residing in your core data center. And that makes sense to have it there. But then you start to move into areas around where you need ultra-high reliability and more ultra-low latency type of things. When you think about it from a manufacturing and robotics, it becomes a key use case where you need to have that low latency. Those things and you start to move at the edge closer to on-prem as an example. And then you've got basically different sensors. And one of the things that you would highlight is an endpoint edge, and that could be one of the things. To kind of cap it all off, though, and as has been brought up is, not only are you identifying where those locations are and what type of specific use cases you're going to have across the edge, but it's really important that you have the ability to have full insights, full visibility and full control in operations across your complete infrastructure. So from your core, your edge, the endpoints, having the ability to have full visibility and full control, full operational capabilities across the entire infrastructure, that's really how you're going to be able to leverage edge infrastructure and then also take advantage of all the efficiencies of deploying multiple different use cases.

So are you suggesting that if an organization does not currently have that type of full visibility across its IT infrastructure even for its centralized data centers that maybe they need to get that into shape before they begin edge expansion?

Yes. It's one of those chicken and egg thing. It definitely helps, right? So you can deploy edge services without the operational. But when we're talking to our customers, we would like to go in and say, here is the infrastructure, here is having the full operations and visibility across the entire network. Because what happens in a lot of cases is you may have some sensors that are relying on your edge infrastructure. And you need to make sure that edge infrastructure is running at peak capacity and that if it does require additional capacity, if it does require additional compute storage networking resources, you have the ability to control it and add that dynamically, right, as opposed to going on site and going there and deploying additional infrastructure there. So by having everything all automated as well as having full operational visibility, it definitely will improve your efficiency in rolling out edge services.

Sure. Yes, that makes sense. Thank you. Eric, how about you? What's your general take on the enterprise edge and where the market sits today?

So right now, the market is early. But at IBM, we see a huge opportunity across the company. In fact, we had an edge-centric launch across all divisions of IBM, not just IBM storage, the first week of May. So we talked about all the different technologies. From a storage perspective, we see the 3 key applications and workloads: big data, analytics on the data at the edge and AI at the edge. So we see that those are the applications. There is also the need to take edge and move that data to the core for further analysis -- analytics, if you've got certain things you're trying to do in AI or, let's say, analytics. So if I am a cell tower company for AT&T or British Telecom or any of the big telcos, that towers all over the place and I create with sensors, an analytic piece of software that tells me Tower 1 is okay or here's what's wrong with Tower 2, you do need to send all that data to core because, overall, you want to find out is that on the tower right? Are there components from my suppliers that aren't working, that seeing 42% of those components fail across the 47,000 towers I have in the State of California as an example. So that you need to take the edge data, which would come right off that cell tower and push it into the core. So those are real things we see that people do with AI, big data and analytic workloads at the edge. And in certain instances, tying that back to deeper analytics, AI or big data workloads that need to take the data from the edge and analyze it from all over the world for sake of argument and understand that data. So that's where we see the edge is. We do see that one of the key things from a storage perspective is tight compact solutions. So we actually announced one of those, actually October of last year that not only work in a core data center, but also could work in a closet. So you mentioned in your presentation, that might just be in a little closet somewhere, where we've got something out fit in a little tiny closet. So we're trying to make sure that we can look at things that we see the edge and extension of the core and the core has the extension of the edge. It's a bidirectional movement of data back and forth and think it's just about the core data centers or just about the edge is the wrong way to view it. You take a holistic approach across a company because while we're all IT guys here, we're not the guys that build the components for the cell phone towers, right? We are not the guys who are doing manufacturing in the auto industry. And so it is very important to take a very holistic view of edge versus core, whereas sometimes IT people see it's either edge or it's core. And that's not how real end users view it based on their applications, workloads and use cases.

Yes. Yes, I think that's a great point. And I think that's how we see the space evolving as well because the infrastructure and successful deployments that we see now, again, very much resembles what's sitting in the centralized data center and not just necessarily on a much smaller scale, but also from a management perspective as well so that everything is or as much as possible is relatable when it goes to the edge and is manageable. And yes, I think if organizations are looking at things on a more disparate level, I think that's where some of the challenges really start to creep in. Let's move to use cases. Misbah, have you seen examples of effective customer use cases at the edge? And I assume you have seen some, which of those have been most effective?

Yes. We've looked at -- across all of our different customers, what are some of the more prominent use cases that are coming up. And as I highlighted earlier, I think there's 4 major edge applications or edge use cases that really justified building out an engine structure. Of course, there may be many more down the future. First of all is around health care, being able to have your health information, being able to do things around whether it's surgery or having the equipment required at your facility, all of those things require the ability to process at the edge and then be able to, at some point, take some of the information that you may not need right away to some of the analytics and then move that over to the core of the data center as well. So we see that very, very tightly coupled. Manufacturing, we see it as another big use case. We see that many customers of ours are using IoT, robotics as manufacturing use cases where you want to have the ultra-high reliability and throughput and low latency at the edge. And we see that with a lot of the robotics and manufacturing use cases. And an interesting one that I focus in on is on the telco side. And the telcos, they're not necessarily building out the use cases, but they are developing the infrastructure to be able to support multiple different edge use cases. And for example, we have multiple customers who are right now, they have the core network and now they're looking at how do we actually take some of our existing infrastructure, move that to the edge and so be able to have that as a software platform, that then application developers can then use that to build out new exciting applications, right? And so this could be your AR, VR applications. It could be other IoT applications. It could be B2B vehicle type of applications. So all of those can then design on a telco platform. And so that's what we're seeing as the -- some of the key use cases and key drivers for edge computing.

Thanks for that. So do you feel that enterprise customers or some of them feel that it may be difficult to engage with telco providers to leverage the benefits of the -- of what they build out with those platforms at the edge, simply because the relationships haven't been there up to this point? And if so, how do they plan to overcome that?

Yes. That's a great question. And telcos, in general, have been -- they have been going through, I wouldn't say necessarily pain points or growth -- growth in pain. But what they are seeing is that for them to be very competitive, for them to be able to kind of attract and still to keep those enterprise applications and enterprise customers on board that they do need to build out that edge infrastructure and that they need to extend that relationship and that infrastructure to the enterprise customers. A perfect example that we're seeing right now, which is around just the magnificent growth of the hyperscalers. And we're seeing where hyperscalers, such as AWS, Azure, they all are coming out with edge strategies. And so for example, just yesterday, Azure announced a partnership with Verizon for deploying edge. And so we're seeing some of that momentum where the telco is marrying up with a hyperscaler to go to a developer. And the relationship there is going to be where the telco can provide the infrastructure and basically the core infrastructure where then the hyperscaler can go in and then start to build out that edge application ecosystem. And so we see that as a really important area and a really compelling area for enterprises to get involved in. And so again, from a VMware perspective, a key aspect of that is providing that common infrastructure across enterprise applications, enterprise infrastructure, core edge, being able to extend those workloads to the public cloud as well as then move out to the telco cloud as well.

Sure. Yes. That makes sense. Sudhir, I'm curious to get your thoughts here. Dell certainly has built out its edge strategy in recent years. Curious to get your take on the use cases you're setting out there in the market.

Yes. We're seeing a lot of great use cases over the last couple of years. I would say -- so we talked about IT closet, right? We talked about having to deploy an IT cloud, and that's an economical example we've heard of for many years in the edge. One interesting on that came up very recently, we're going to be hopefully announcing it soon is, it's actually running at the home station for your entertainment, right. So when you go to an entertainment park -- theme park, and you go in the roller coaster and come back, right. When your car docks into the base station, it's transmitting a whole bunch of data about all of the mechanics and the sensors -- using sensors along the tracks, and that's all being said into this analytics engine that's powered by our software running on our gateways right there at the location. So I mean that's how distributed it's starting to get. And so using that, they're able to detect whether this ride is safe to go again or do they need to do preventative maintenance? So it's really changing the definition of where this thing gets deployed. Another one similar to that, obviously, is airports, right? The plane comes in, lands, pulls into the gate, terabytes of data coming off of that plane. Again, for the same reason, they want to try to quickly assess if there's any reason for the plane to not go on with the next journey, which is probably 45, 50 minutes away. So that whole notion of real-time -- there's no time to send the data to the cloud or anywhere else. It has been to be done right there. We're seeing, obviously, manufacturing, both from a -- is your manufacturing floor running properly, efficiently, but also now getting into the quality of the products that are being produced in the manufacturing in the pipeline, right? This visual -- using computer vision to detect anomalies in the product and then try to catch it early in the line. We've actually talked publicly about some of our agriculture use cases where these farms -- indoor farms that are all highly instrumented for regulating the temperature to ensure optimal growth of the produce. So we're just seeing this all over the place. There's so many connected cars, transportation, talk about health care. That's also certainly very big where we have use cases where these small devices are going to people's homes, right? The health care providers are taking them to the home to record the vital statistics of the patient that they're caring for at their house. And that's all being analyzed right there to provide instant analysis, diagnosis or whatever, but also being moved back for further long-term analysis.

Yes. So I see -- I'm hearing plenty of boosts and efficiency and optimization. And when I hear about edge use cases and associated IoT use cases, I tend to hear about things that many customers haven't even thought of before in terms of improving things along their supply chain or on the manufacturing floor or they thought it was never possible, but it seems to me that edge computing is really enabling a lot of things that were just previously unattainable, which is exciting. Eric, any thoughts from your end on use cases or unique use cases you're seeing among your customers?

So we are seeing use cases that traverse all sorts of industries. Sudhir gave you one, the example of the telco because that's a real example of one of our customers who's using that data from the edge back to the core, analyzing that data. We have another customer, and this led to what was brought up already in the manufacturing space. In this case, they're actually a bottler. And so they're doing analytics and quality control at the end of the line. So is the cap on right? Does it look like the cap was messed with? Then also quality control going into the packaging, which is, as you know, first of all, could be something like a 6-pack, then 6-pack, goes in a bigger box, it then fits on a pallet. And then, of course, you have to pallet wrap so that it can go on to either a truck, a boat or a train. So they're doing really in this end -- for that company, they're doing at the end of the line. They're not doing at the beginning of the line. They're doing quality control on everything from the packaging side, in this case, literally getting it ready to go into that truck, but also on the actual bottle itself and the cap itself. As you know, there are some serious issues with people getting into things in the early 2000s, they were drugging Tylenol and people were doing stuff. And you can see that in the packaging. If you just go to -- when you buy your gas, you pick up a diet soda or something or certainly if you do pick up something like a small milk, right, they're in a plastic ball, you open the top and then there's another safety vehicle in between. So they're doing that kind of applications workloads and use cases. And then clearly, we see the same that Sudhir mentioned on the medical side where people are trying to process data, but it's got to get back. Because if it registers the data, unless it's a critical situation, obviously, to call an ambulance. But Sudhir and I are not doctors, right? And so far wiser, kids are sick, it's got to go to somewhere and then that doctor or nurse has got to be analyzing that. And if there is a bad situation, getting back to us because Sudhir and I, we don't know what to do, right? And that's part of, again, the edge versus the core. The other thing we're seeing just from a practical perspective is on the edge, you need to make sure you've got the high availability and the high reliability. Because if we can do something like the health care monitoring or, in this case, the production line monitoring at the end, if the storage of the servers, the infrastructure around the actual software goes down, there is no bottles coming off the end of the line, right? You can't because it's all regulated, right? Health care is heavily regulated. The food industry is regulated. So in all -- the airline industry, Sudhir brought up. All these regulated industries, you've got to make sure the edge talks to the core, and you also got to make sure the edge doesn't go down. That doesn't mean a sensor might be bad and companies like -- that are on this panel, we don't control that. What we certainly control the infrastructure that, that sensor device is talking to, the storage, the servers, the software infrastructure in the case of VMware. So if those things ever go down, now you're the bottleneck. Same thing in performance, you don't want to be the bottleneck either. So your performance needs to be highly scalable. We have, for some of our things, our product once called Spectrum Scale. We have documented end users where they're getting 2.5 terabytes a second of bandwidth in the real world. Well, for big data, analytics and AI, that's what you need. Now it doesn't mean you're going to get that at the edge. But again, the edge center to the core and the core has got to analyze it quickly and get back. Health care is appropriate example like the home monitoring that Sudhir mentioned. If they're not analyzing that quickly, some could die, right? So these are critical situations that if that edge device or it's the same thing in the telcos, right, or obviously, there's been issues in California fire. So if all those PG&E towers or any of the other electrical providers across the country, this is antiquated infrastructure, auto, same thing, right? The roads, the bridges, all of these guys are getting into looking at what they've got. So if they're already having some problems for sake of our example like PG&E has publicly talked about because they've obviously been in trouble with the government, both state and federal, then you can't be the one that's the bottleneck. You can't be the one that doesn't send the data because your storage array or your server failed at the point of attack so the data doesn't go back. I think Sudhir's example of the airline is like certainly a real-time one. When that thing comes in, I see [indiscernible] doesn't get it right. I mean, that's even worse than something like a cell tower problem or the power line problems. So that's what we see, and it's becoming very ubiquitous, right? And people don't realize there are things that [Audio Gap] one and all of this is being done by computer and one person versus hundreds of people like in some of the old movies, you see all these people at the end of line sorting, like there's a famous written one in that, maybe show I love Lucy. Well, guess what? Now you don't have that. All that chomp is being done automatically, you don't have that funny scene anymore. But you better hope you support your servers and your VMware doesn't fail.

Yes. Yes. That makes sense. Thanks for that, Eric. So we have about a minute left for each of you. And I'd like to ask a two-pronged question. So just spend about a minute, I promise not to mute your mic like the moderator is going to do on Thursday night. But just spend about a minute. Misbah, we'll start with you. What is VMware doing to help its customers with the edge? And what's it doing to prevent some of the barriers that I talked about?

So yes, within a minute, essentially what VMware is doing is we're providing the infrastructure, right? And so that is the software infrastructure, that will allow you to enable the enterprise customers to build out applications on top of that platform. And so one of the things you mentioned, Christian, is that the use of Kubernetes and how that's actually going to accelerate time to deploy, it's going to accelerate kind of application development. And so with our infrastructure, we're adding in a layer that we're calling Tanzu, which is an application platform that allows enterprise customers to build applications very quickly regardless of if it's at the edge, core, or even in the cloud. And then I think the most important aspect of this as well and between Eric and Sudhir, they've all kind of agreed on, which is having that control, having visibility so that you do understand exactly what is happening at the edge. If you need to add resources, if you need to look at the health of that infrastructure in place that you're able to monitor it and then also be able to take remediation capabilities to actually solve and then inhale those issues. So that's from a VMware perspective, being able to have that infrastructure, application deployment on top of that and then management and operations to that level.

Great. Thanks for that. Sudhir, I'll go to you. What is Dell doing to enable the enterprise edge and overcome some of those barriers? One minute. Sudhir, you might be muted.

I am muted. Sorry about that. I'm going to try to keep lost 10 seconds right there. So Dell is doing 2 things on the hardware side and then the software side. On the hardware side, we've got servers and gateways that are edge -- designed for the edge. We have a number of announcements there. But we also have storage appliances like power store that can be both a pure-play storage device or, like you said, have the ability to run applications and compute on the device, what we call apps on. And that was recently launched, and that's perfect for that edge deployment, along with our other hyperconverged products. But I think the more important thing is on the software side, we've got the best portfolio of software-defined block storage, object storage and now we've also introduced a new software-defined platform called the streaming data platform. We're doing real-time analytics right at the edge. And in terms of helping customers overcome, I think it's providing that end to end solutions so that they can see the ROI from some of these investments due to operational simplicity.

Excellent. Thanks, Sudhir. And Eric, and closing thoughts on how IBM is helping to enable the enterprise edge.

So first of all is always high availability and reliability and performance. You need that at the edge. You need it to process. You need it to make sure it's always up and going. I had a number of examples discussed that could be critical in nature, such as the health care, the airline one, others that are just huge financial impact like on the cell tower example or in the example of a manufacturing line. The other thing is automation, automation, automation. There isn't an IT person or a storage person or a server person or VMware admin usually at the edge. So you've got to point and click, set it and forget it. They may check on it once a while, but the bottom line is, it's got to be unmanned. It's an unmanned plane. There is no man. There is no women. There is no anything. So you've got to make it easy to go there. And then the last thing, of course, is remote management and control. So you've got guys in the data center [Audio Gap] right? In this case, it would be more than a NOC and you'd be able to go to any edge device, in this case, storage in our case. To go to the edge, you'd be able to reconfigure the array, see if there's something wrong with it. So those are the things that we're doing to try to keep these systems up and going at the edge and to make sure that they have the right feature sets for an edge deployment.

Great. Thanks for that, Eric. And that's all the time we have for today. I'd like to sincerely thank our panelists for joining today, and also for all of you for taking time today to hear our thoughts on the enterprise edge. We do hope you found this session helpful. And with that, I'll turn it back over to Eric Hanselman.

Thanks, Christian. This closes out our Fifth Day of 451 Research Hosting and Cloud Transformation Summit. We were really on edge today. And I hope you had a chance to catch all the sessions. To recap, Christian Renaud spent time exploring the edge potential, which is especially important given that the voice of the enterprise response, identified that more than half of their IoT data is initially stored and processed at the edge. All of that edge was tied back to data centers by Craig Matsumoto's session on edge interconnect. While many enterprises are still taking an old school view of how to leverage more dynamic data pads, there's a lot of untapped potential that Craig pointed out. Brian Partridge's panel looked at today's realities and the future potential of the telco edge. It turns out there's a lot of value beyond the hype of 5G, and organizations have to be ready to put it to work. And as you've just heard from Christian Perry and his panelists, enterprise edge has its own dynamics. As gateway functionality gives way to fully functional edge computing, but some education is required for enterprises to capitalize on its full potential. If you missed any of this, these sessions and all of the last 2 weeks are available on replay in the auditorium. We hope that you got your calendar marked for our final day on Thursday, too. We jump back from the edge to explore customer experience and digital transformation, including data-driven experiences and cloud-based payment systems. Our HCTS Chair, Research Director, Melanie Posey, is the host for our final day and will be leading us into some intriguing topics. We close out HCTS with a 451 analyst wrap up panel that looks at the future with a discussion of how to deliver better business outcomes, levering IT, of course, in the digital era. Please also take time to complete the feedback surveys for today. This is our first year going virtual, and we want to get your feedback to ensure that we're getting all of you all the information that you need. We want to thank all of our sponsors who've made all of this possible. I hope that you've had a great day, and we hope to see you on Thursday. Thanks for being here.