Microsoft & Dell: Powering the Future with AI and Hybrid Cloud | BRKFP302H
[MUSIC] Bernardo Caldas: Hi, everybody. Good afternoon. Welcome to our session on the Dell and Microsoft partnership, powering the future of AI and Hybrid Cloud. I'm going to kick things off today, and sharing stage with me is Varun.
So how about you introduce yourself? Varun Chhabra: Sure, thanks. Bernardo Caldas: [inaudible] . Varun Chhabra: Thanks, Bernardo.
Hi, everybody. Varun Chhabra, I work at Dell in the Product Marketing space, and I'm responsible for our infrastructure solutions as well as telecom marketing. Really excited to be here. Bernardo, do you want to introduce yourself? Bernardo Caldas: Yes, and I'm Bernardo Caldas, I'm Corporate Vice President, responsible for our Azure Edge Product Management, and been really a key part of this partnership that we've had over many years. That's where I wanted to start our session today, talking about all the great work we have done over the years. I've been at Microsoft a long time.
We have a 30-year-plus partnership. I think I've been part of that journey for about half of that, which is still pretty impressive. What's cool about that is that we've been able to see changes in the industry and spot opportunities together. Varun Chhabra: Yes. Bernardo Caldas: Then really bring to bear the expertise from both companies, to solve problems that we see from our joint customers.
What you're going to see today is a prime example of that, where we're really innovating in the area of Hybrid Cloud, bringing choice, flexibility, and really looking into the future with new opportunities such as GenAI coming around, where we're able to bring these experiences to the Edge, or really anywhere our customers are looking for. Varun Chhabra: That's right. Bernardo Caldas: It's really great to be doing this session together. I'm going to turn to you to kick things off, and then I'll go from there.
Varun Chhabra: All right. Thanks, Bernardo. Let's start with what we're hearing from all of you. It really comes down to what's the state- of-the-union when it comes to whether you call it multicloud, you call it hybrid, Microsoft is also introducing a term like adaptive Cloud these days.
What we're hearing from when we speak to many of you, whether it's on our side or Microsoft, we do a lot of joint briefings, and there are some consistent themes that come out when we're talking about the pain points that we're hearing from customers. When customers look across their Cloud estate, whether it's different Cloud providers, different applications and workloads running in different Cloud providers, on-premise state, out of the Edge, there is a desire for infrastructure elasticity. There's a desire for making sure that there is a consistent compliance and data sovereignty regime.
There's a need to react to new opportunities in an agile manner. Bernardo mentioned Edge. There's all kinds of things that are happening with businesses, that all of you are asked to react to, and react in an agile manner to support the core business. Then finally, the two-letter word that's on everybody's mind, AI. How does all of this play in with deploying AI workloads? Where your data is? Data is distributed everywhere and you're looking to understand, how do I make sense of my Hybrid Cloud estate in the context of deploying some of these AI workloads that are being asked for by the business. These are very easy- to-understand needs.
But the challenges associated with them really rise from the distributed nature and the siloed nature of Hybrid Cloud environments. The tools that your teams are using to manage infrastructure in one location or applications in one location versus applications and infrastructure in another location, whether it's different public Cloud providers in the Azure environment, on-prem, out at the Edge, they're all different. You've inherited them, different environments, as a result of acquisitions. Maybe there's some shadow IT work that you've incorporated back into taking control of. But there's just a disparate set of tools.
It's difficult to train your team members because there's always new environments and the tools are different. You have team members who understand one part of the environment but they're not really trained to understand and deploy or manage the other parts of the environment because the tools are not that simple. Data residency and regulatory environments, it's changing rapidly. We just saw the Biden Administration issue some high-level guidelines around AI. Of course, Europe is doing that, other countries are going to follow suit as well. How do you make sense of all of that? Where do you deploy your data? Where do you keep your data and stay on top of whether it's local, geographic residency requirements or industry-specific residency requirements.
It can be really complicated as you think about your Hybrid Cloud estate. You have to do all of this and right-size your infrastructure. Your budgets are probably not increasing. They're either staying flat or decreasing, and you are trying to make sense of this complexity. Now, what we hear from all of this, what comes out is, you want solutions that are simple, you want tools that work for your team regardless of where the location is.
If you're running something in Azure or if you're running something on-prem, you want the same sets of tools for your teams so that you can simplify things, standardized processes, and really focus on the most important things. Dell has been on a journey to help customers with multicloud and some of these challenges, and we have a set of solutions that are under the APEX brand. Really, our goal is to help customers across different scenarios in this hybrid and multicloud world. Whether it's taking your on-premise environments and toolsets and using the same technology in the cloud, we call that ground-to-cloud, or taking experiences in Azure and bringing them on-prem, we call that cloud-to-ground, making sense of that entire estate.
How do you do that with consistent toolsets and with minimal impact to your team, so that they can take their existing skill sets and bring them to bear. That's really what our goal is. The reason why Bernardo and I are so excited to talk to you today is, how all of this is coming together is in a few key solutions, the most important of which is the APEX Cloud platform for Microsoft Azure. What we've done here is, taken what is the Azure Stack experience, and sat down and looked at customer feedback and thought about, how do we simplify deployment of Azure Stack and Azure Arc services on-prem? How do you do that in a way, in a standardized platform, and keep it really simple? What we've done with the APEX Cloud platform for Azure is that, it's an on-prem solution, on-prem infrastructure that has standardized infrastructure parts. Our Power Edge Compute software- defined storage that's standard across the board as well as MNO software that is co-developed and integrated with Azure Stack, specifically meant to simplify deployment. We call that the Cloud Platform Foundation Software.
The Cloud Platform Foundation Software makes sure that the infrastructure and the hardware is talking to the Azure Stack in a way that simplifies lifecycle management, simplifies deployment. Our current internal calculations show that, when you deploy an environment with this platform and compare that to other solutions for Azure Stack that are out there, you have 88 percent less steps. That means that there is less time spent by your teams to deploy this and more time spent on actually supporting the workloads that you're running on your Azure Stack and Arc environments.
There is a really, really high degree of joint engineering between Microsoft and Dell on this. This is not just OEM software that's deployed on hardware. That joint engineering means that we are working together to test and validate our solutions and the code continuously. So much so that whenever Microsoft issues updates to any part of the Azure Stack software or other pieces of the Stack, Dell will issue a corresponding update within four hours of Microsoft issuing that update. Which means that, if you're using this platform, you can be sure that, not only will the Microsoft updates be available to you, but you will be able to run those updates and do the lifecycle management in a tested and validated state right after Microsoft issues the updates on their side as well. So there's a lot of close integration.
It's a great way to take the Azure experience and bring that on-prem, or take your existing on-prem Windows experience and take advantage of that. There's integration both with the Azure Portal as well as with the Windows Admin Center that your teams are using today. It's built to really be managed by your existing tools, have your team members be able to use their existing skill sets and not have to spend too much time learning something new.
Because of this reason, the tight integration and the strong core development that's happening, we're actually the first solution that's going to be in Microsoft's premier category of HCI solutions. It's really a sign of the core development and the strong integration work that's happened. That's a little bit about the Cloud platform.
I'm going to ask Bernardo to come in and talk to us about the value that the Azure Stack and Arc and other pieces of the Azure piece bring as part of this joint solution. Bernardo Caldas: All right. Thank you, Varun. I'm terribly excited about this partnership that we've had because it was really months of work that led to the launch of the APEX Cloud Platform for Azure. Really, that tight integration, that tight partnership, and then really looking across what are the big problems that we need to go and solve for our customers, and then bringing a solution to market, which is now available and ready for you to go and deploy. The solution is built on Azure Stack HCI, as Varun said. Azure Stack HCI was an infrastructure that we created for customers that are looking for a solution to bridge their on-premises deployments in the Cloud.
When we go and talk to customers, we see the majority of our customers have a mix of on-premises deployments and Cloud deployments, but they're looking for a way to consistently manage that across their entire estate. What we've done is we've backed on a set of technologies, which I'll describe later around Azure Arc, to do things like at-scale management through the Azure Portal. Then being able to run, whether it's traditional VMs as well as Cloud- native workloads, on top of a Kubernetes solution that's part of the platform and available. So you can do unified lifecycle management, as Varun said. When you do an update, you can do a full stack update to minimize downtime, minimize risk of solutions going down. This whole thing is delivered as a service.
The business model that you are familiar with with Azure is now available for this on-premises infrastructure, so you can consume it as a service as part of your Azure commitment, for example. In the latest update that we announced here at Ignite, we call it Azure Stack HCI 23H2, we did a bunch of work to really integrate with Azure Arc by default. That means that everything that you need to connect that solution to Azure, it's available right as you deploy it. It's really also transcend even to the support experience.
When you think about having stayed in Azure and on-premises, you have a consistent support experience across both of them. We also offer flexible deployment options. You can deploy single node.
This is a question I've got a lot in my customer meetings during Ignite. How small can you go? You can start with a single node and then you can add up to 16 nodes per cluster. With the capabilities that we've built through the Azure Arc integration with the Azure Portal, we can also do fleet management. You can manage multiple of these clusters. It could be big clusters in a few locations, or a lot of small clusters in thousands of locations.
All of that is enabled. Of course, one of the drivers for customers that want to deploy this kind of infrastructure, is being able to operate when you're not able to connect to the Cloud on a permanent basis. It could be, you lose a connection and you have to continue doing business. That's fully supported in the sense that we can support intermittent connectivity with these solutions, which is really powerful for many of the use cases that you care about. We see this really popular in industries like manufacturing and retail. We're seeing credible momentum.
People really excited about trying this out. Really, this whole experience builds on top of a much broader story, which is the Azure Arc story. I'll spend a few minutes talking about that as well. Today we have about 21,000 customers that have deployed Azure Arc in their environment. That number grows every month.
It's really exciting to see the momentum. The idea behind Azure Arc is to extend the same experience that customers have in the Cloud to multiple on-premises environments. That could be different Clouds, it could be enterprise core data centers, it could be the Edge.
Multiple environments that they have, they can use the same dev tools. They can use the same data tools, the same AI tools that they have available in Azure. Even the Copilot experience for Microsoft Azure has now been extended to support Arc-enabled resources. Just imagine you can enable all these different environments across that and use Copilot to reason against that, and that's a new capability that we announced here at Ignite.
Double clicking a little bit into how this platform works. We start with the control plane. The Azure control plane is, of course, built to handle the load of your entire Azure estate. It's one control plane that control all the resources in Azure. It's a pretty extensible control plane.
That's the starting point for this experience. Because what we do with Azure Arc is, you project resources there on-premise to that control plane so you can then use Azure Services with those resources, whether they're management, security, identity, monitoring. All those become available the moment you bring an infrastructure up with Azure Arc.
You're able then to use modern applications and data. That includes past services. There are Azure Hyperscale Services that are coming on-premises on top of a Kubernetes platform. Here we backed again on industry standards.
It could be any Kubernetes platform. We have our own with Azure Kubernetes service. But this can work on any CNCF-certified Kubernetes solution, which then allows you a lot of flexibility in what you can deploy. This works on any environment including Azure Stack HCI, which is the one I just talked about.
Now, what's unique about Azure Stock HCI is this whole thing comes pre-integrated. Varun mentioned a few times, simplicity, time-to-value, cost, complexity. We've removed all that out of the equation with Azure Stack HCI, and we bring this whole stack ready for you to go and deploy on your environment.
This next slide, it's a bit of an eye chart and it's not even everything that's available, but gives you a sense of what this technology stack looks like. At the bottom, we have the operating system that powers all of this, including Hyper-V, Storage Spaces Direct and Software- defined Networking. It's a hyperconverged infrastructure for resiliency.
We do start at single- node and then you can grow from there and add more nodes as you need. The platform is able to run virtualized applications, and that could include Windows VMs, Linux VMs, and even Azure Virtual Desktop, which is also coming soon in the platform. But many customers are also looking to take a platform that can handle their legacy applications, but also be able to handle the new applications they're developing around Kubernetes using Microsoft services. That's what we have on the right side there with the Azure Kubernetes Service hybrid that runs on the platform.
Again, in this latest update, this comes pre-integrated. When you deploy the latest update of Azure Stack HCI, it already comes with all of this. Then with that, we're able to run Azure services on top, whether it's machine learning, data services, app services, and more. We're growing the catalog of services each day. In fact, at Ignite, here we announced Azure IoT Operations, which is the newest Arc- enabled service that's used to connect data from sensors in operational settings, up to Fabric for data processing and analytics. The power of this is that you can use all this Edge infrastructure in conjunction with the public Cloud.
What you see at the top there are popular services that our customers use along with the solution, whether they're management and security services like Defender, Policy, Update Manager and so on, as well as other past services there in Azure. Most of our customers are looking at solutions that cut across Cloud and Edge so you can have part of the solution running the Cloud, part of the solution running at the Edge, and the combination of the two is what's really powerful. Of course, the work that we've done together with Dell, with the APEX Cloud platform for Azure, really bring this all together with a whole package.
We couldn't make it easier for you and more robust for you. We collaborate very closely from an engineering perspective. We have Dell hardware running in our labs where we do integration testing. The commitment on both sides that the solution is the best that you can get with this platform. With that I'll turn to Varun.
Varun Chhabra: Thanks, Bernardo. We talked a little bit about the APEX Cloud platform. This is about extending Cloud experiences, Azure experiences, to on-prem. Similar tool sets. Now what I want to do a little bit of is talk a little bit about what Dell has that are helping customers take their on-premises experiences with Dell software and extending that into Azure. This is what I was saying is basically ground-to-Cloud.
Let's look a little bit deeper into that. One of the things that Dell is known for is really helping customers with their data, whether it's data storage or data protection. The reason why we have been successful in helping customers with this is, we have unique IP in our storage software and data protection software that customers have come to rely on for their most demanding workloads, achieving scale, running their mission- critical workloads on our storage platforms. What we're doing with APEX is also taking some of that IP and making that available on Azure for you to support your Azure workloads in the Cloud with as well. We have something called APEX Block Storage for Azure. This is taking our block storage software IP that customers run massive number of nodes on with on-prem, mission-critical workloads and making that software available in Azure to be able to run.
The APEX Block Storage for Azure is able to support extremely high scale, your most demanding workloads that require the lowest latency, your databases, point- of-system records, and all of the most important and demanding workloads that you have for block storage. This platform can manage at scale because you will be using the same software across on-prem and Azure. You can take advantage of the same MNO toolsets for managing your data across different environments. This is a huge ask from customers whenever we speak to them.
If you think about not having to re-platform your block storage applications as you're moving them from one location to another, this can be another huge source of savings and efficiency and agility. What this then does is, as you look at getting one view of your data across different locations, which, especially in the AI world, is becoming more and more important. Having that unified view of your data, no matter where it's located, whether it's in Azure, in your data centers, or on the Edge, this can really help with that. One of the nice things about this is that since so much of what we deliver to customers in the storage space is mission-critical data protection, making sure that the applications don't go down. What we have available as part of APEX Block Storage, just an example is, the storage software actually works natively in a multi-AZ environment to make sure that there's multi-AZ availability for your data without having to do anything extra in terms of snapshots or replicating data. The platform just does that for you natively.
It's a huge source of savings, so if you're deploying this across multi-availability zones and turn on the setting around data being striped across these AZs, the platform will negatively manage that for you without you having to pay extra for other snapshots, etc. This is a huge savings, huge DCO savings right from there. Since this is available on the Azure Marketplace, you can use any committed spend that you have with Azure, with Microsoft, for this software. This is a huge area that I would ask you to look into as well if you haven't already done that.
This is storage. We have the same thing with data protection. Data protection, we have a massive install base for data protection, and one of the things that people are always surprised about is, our customers are using our data protection software to actually store and protect data in the Cloud as well. Whether you're trying to backup your on-prem workloads data to the Cloud, or you're trying to protect data for your applications in the Cloud already, protect that data, or set up Cloud disaster recovery options for your data centers or on-prem infrastructure and workloads, we can help you with that. This is not an area that's new for us. We actually have over 18 exabytes of data that customers are protecting in the Cloud today.
Every time I do this presentation, we have to update that number just because it's just a massive data estate that we're helping customers within the Cloud. We have over 1,800 customers that are using this protection software to protect their data in public Cloud. Just like I said with the previous one, available on Azure Marketplace so you can use your committed Azure spend for this as well. Then finally, what many people know Dell for and Bernardo was talking about the Dell and Microsoft heritage going back to 35 years. It started in the PC space, and that of course continues to be super important. With all the Copilot announcements that have happened this week, I would be remiss if I didn't mention that as an area where Dell and Microsoft are collaborating with as well.
We have our PCs. We work very closely with Microsoft to make sure that our PCs are the most secure and most reliable PCs for your end users. As you are looking to roll out Copilot and you're looking at upgrading your PCs for your end users, we're doing a lot of integration work with Dell to make sure that that is a great experience for your end users.
Then finally, for those of you that are looking at actually deploying a Copilot in your environments, we also have professional services that will help you with assessing what work needs to be done with your data, safeguarding, etc., consulting and deployment help with the various Copilot services. Any help that you need from an implementation perspective, Dell and Microsoft are working together to help you with that as well. With that, I would just say, whether it's in the Cloud space where bringing your Azure experiences on-prem with the APEX Cloud platform or taking some of the storage and data protection IP that you may be taking advantage of from Dell on-prem and taking it into your Azure environments, or on the PC side where you're empowering your end users, there's a whole host of collaboration that Microsoft and Dell are working together with. We would love for you to engage with us. There's a QR code here.
There is a video that Bernardo did with one of my peers at Dell, talking a little bit more about the APEX Cloud platform and going a little deeper into what makes that so special and why that's such a unique collaboration for the two companies. You're more than welcome to check that out. We also have a booth where you can learn more about the platform and some of the other storage software and data protection software we talked about today. Really, really appreciate you coming here today. We're more than happy to take questions from the audience or online as well. (applause) Anything online, Mike? Bernardo Caldas: Stunned silence.
Varun Chhabra: Stunned silence. Yeah. We answered everybody's questions. It's a post-lunch session, everybody is just working on their lunch, digesting their lunch. Yes? SPEAKER 1: Is there any factory-integrated option available from the APEX Cloud platform? Varun Chhabra: It does, actually. The question is, "Are the options we're talking about here like the APEX Cloud platform, is that factory integrated?" Absolutely, the software will ship from the factory.
As Bernardo mentioned, the testing and validation on the Cloud Foundation software and the testing with Azure Stack is done all beforehand so that you don't have to actually manually deploy that on-prem. Mike, anything to add there in terms of factory integration? Mike: No. You might have to be ready to just give the microphone in order to be heard on the web. Varun Chhabra: I'll repeat the question if that comes out that way, yes.
Thank you for the question. There's a question online that says, "How is this licensed? Will clients be able to leverage their current EAs or contracts for Fed? Mike, do you know the answer to that? MIKE: No. The Fed, SLED, and LATAM are coming very soon in Q1. SPEAKER 2: That micro will not be going to broadcast, so you'll have to answer the question.
Varun Chhabra: I guess I'll do it. Specifically for Fed and state and local government and LATAM, that solution will be coming soon. But in general, with your current enterprising agreements, I'm assuming you mean Dell, I'm not sure if you mean Microsoft. Bernardo Caldas: I can talk to the Microsoft one. Varun Chhabra: The solutions we talked about today, you can add them to your ELAs for Dell, and Bernardo, maybe you can talk about the Microsoft. Bernardo Caldas: On the Microsoft side, so the business model here is usually an Azure subscription, so you pay for the consumption.
If you have an enterprise agreement, a Windows Server Software Assurance Agreement with us, we have a program that allows you to exchange essentially software assurance to Azure Stack HCI, and you get access to Azure Kubernetes Services as well. If you have an existing EA or SA with us on Windows Server, you are able to use that with your Stack HCI solution. Varun Chhabra: In short, whether you're looking at your Dell enterprise agreement or Microsoft, we have ways to integrate that into it, absolutely. SPEAKER 3: In regards to the use of Copilot with APEX, I understand on the Microsoft side, the capabilities coming and so forth, is the vision to allow Copilot to give you the ability to interface with the hardware as well, or is that going to be abstractive? Varun Chhabra: Great question. So question is about Copilot and how it integrates with the Dell hardware.
Right now, we've not announced any plans yet on that space, so stay tuned. Given that some of these announcements came out just this week, we'll have to work with Microsoft to figure out what that roadmap looks like. Bernardo Caldas: We're in preview with some cool scenarios around anomaly detection, for instance, in your HCI instances, so if you go and look at the documentation we put out for the preview of the Copilot for Azure, there are some Azure Stack HCI scenarios there, they are available for you to try.
It's still in preview, we're still working through. But the possibilities are endless, really. We're just scratching the surface with what's possible.
You've seen all the innovation of Copilots throughout the show. You can only expect that more of that will be available in these hybrid solutions that cut Cloud and Edge. Varun Chhabra: Given that we're running the Azure Stack on the Cloud platform, anything that's integrated through Arc or Azure Stack should technically be things that customers can take advantage of. Yes. Another question online from Kishor.
"Do you have powerful GPUs on the laptops to run the workloads or inference on the Edge?" Answer is absolutely yes. The one thing that sometimes gets lost in all the GenAI hubbub is that, this is not just a data center conversation. People are increasingly asking us for support to run and tune GenAI models on their devices as well. We have a whole range of GPU- supported or GPU-enabled laptops and workstations that customers are already using to start to tune, and in some cases, even train smaller models.
For example, we made an announcement with Facebook, I think it was two weeks ago, around the work we're doing with them on LLaMA 2. We actually have a demo of somebody running a LLaMA 2, I think seven billion parameter model, and tuning it on a Dell workstation. That's already possible today. Hopefully that answers your question, Kishor. Yes, please.
SPEAKER 4: When we talk about APEX Cloud platform, are we referring to the Dell Cloud platform? If it's yes, then how does it indicate with the existing Azure Microsoft Cloud platform? What advantage that it brings to non-existing users? Varun Chhabra: Sure. Thank you for that question. The question is, the APEX Cloud platform for Azure, if it's built on Dell infrastructure, how does that integrate with the broader Azure services that customers have? Did I get that question right? Essentially the answer is, absolutely yes, that's what it is designed to do. What we've done is, because the system is based on Azure Stack HCI and built to support Azure Arc services, the design goal has been to actually support anything that you're running on Azure that's supported through Arc or Azure Stack HCI, bringing that on-prem, and having that consistency on-prem.
The work that Dell and Microsoft have done is to make that on-prem deployment and the on-prem lifecycle management, like updates and things like that, as smooth as possible. Really trying to provide that best-in-class Azure Stack HCI and Azure Arc experience on-prem. Bernardo Caldas: Yes, that's exactly right. The way to think about it is all built-in in a sense that it can come pre-installed, it comes pre-installed from the factory.
With this latest update as I mentioned, it even comes with Azure Kubernetes Services also built in, so that whole stack that I talked about, that's included in the solution. Then you can then consume Azure services from that point. The idea is you turn it on, and then the solution connects up to the Azure Cloud with very few steps. Then from that point, you have a Cloud endpoint or Edge endpoint that's managed through the Cloud that you're able to deploy applications to, and services.
Varun Chhabra: Great question. SPEAKER 5: [inaudible]. Varun Chhabra: Do you mind if I just go there and then I'll come back to you, sir? Go ahead. SPEAKER 6: Coming back to the hardware and GPU, when I asked portfolio tools, it was mentioned that I can only use one specific graphic unit before because it only works with Microsoft.
Is there anything else coming up, a new feature? Varun Chhabra: Your question is about what GPUs are supported. Are you talking about on the devices like laptops and workstation side or on the data center side? SPEAKER 6: On the working stations. Varun Chhabra: On the workstations. I'll have to get that answer for you, actually. Am not the expert on the workstation side. I don't know if any of the Dell team has that answer.
My guess is there will be a support for a wide variety of GPUs long-term. Because this is, as you can imagine, a rapidly evolving space. Right now there's a lot of work being done to integrate and make available as wide of a GPU ecosystem as we can for different workloads. Do you want to go back to your follow-up? SPEAKER 5: It's a hardware, right? Varun Chhabra: Yes.
SPEAKER 5: Microsoft use Dell hardware, so they're like claim partner. How long have they been claim partner? Microsoft has operating system like Windows 11 and up. They use their own...How does the compactability match the hardware--? Varun Chhabra: Your question is, we talked about Azure Stack, but operating systems like Windows Server etc.,
how does that integrate into this? Same thing. Look, I think the collaboration on the APEX Cloud platform is relatively new, but the 35 years of integration or partnership between Dell and Microsoft has actually been on the operating system space, whether it's in laptops or on the Windows Server side, Dell was one of the first server manufacturers that collaborated in Microsoft and Windows Server as well. I don't think that's anything new. That's something that the companies have a lot of muscle around long-term. I don't know if I answered your question. SPEAKER 5: Yes.
That's our bread and butter, that's what we've been known for in terms of partnership. This is great. Well, we really appreciate all of you taking the time and coming in, and love the questions that came in. Thank you so much for your time. Bernardo Caldas: We will stick around, if you have more questions you want to come by and ask, we'll be happy to answer. Varun Chhabra: Happy do that.
Bernardo Caldas: Thank you. Varun Chhabra: Thank you for your time. (applause)