AWS re:Invent 2022 - AWS wherever you need it

Please welcome Vice President Storage, Hybrid Edge and Data Services, AWS Werner Vogels.

Hello everybody. Welcome to the show. Welcome to this conversation we have today on everything hybrid edge. I'm really excited to be part of this conversation today and to help tell the story on what AWS is doing from the regions, from the cloud all the way out to the evolving edge.

I'll be joined by my colleague and partner in everything hybrid edge, Jan Hofmeyer, who's the VP of EC2 Edge. And I'll be joined by Technical Program Manager, Principal Technical Program Manager, Swami Sivasubramanian. And the thought partner and customer, Senior Vice President SE, Sean Sweeney, at Fanduel.

Now, Priyanka and Shane, when they come on stage later, are going to be bringing into focus all of the ideas and thoughts that Jan and I will talk about today. And they're going to talk about how Fanduel is bringing an amazing customer experience to their customers using the full extent of AWS continuum of cloud to hybrid edge.

Okay. So let's get started today. Customers want to use AWS services in a growing range of applications operating wherever they want whenever they require. And they're striving to do so to deliver the best possible customer experience they can, regardless of where their customers or users happen to be located.

One way AWS helps customers accomplish this is by bringing the AWS value to our regions, to metro areas, to on premise and to the furthest evolving edge.

Now, AWS services at the edge help Prime Video deliver new television shows like Peripheral to anywhere on the planet. Now personally, this is one of my new favorites and I wait for it to drop every Friday night. So I'm pretty happy that we're doing this.

They help gamers achieve ultra low latency play regardless of where they are. And for my two boys, this is the one thing that they enjoy that, you know, their father is helping do so. They love gameplay.

We help deliver 5G to applications to stadiums and factories. And when an emergency arises to quickly respond and establish life saving services, communications and support in disaster relief areas.

They will help Nasdaq seamlessly transact hundreds of billions of dollars in transactions for hundreds of millions of customers around the globe.

They enable us to collect, process and make real time decisions on data in the depths of space and in more earthly pursuits, whether from air, sea or land back on the roadways here on terra firma.

Now, in short, we bring you AWS value, its services, its APIs, its tools, its operational practices to wherever you need it.

Now, we've been providing some of these hybrid edge services for, for many years and it's just a personal factoid. I've been with AWS for now 10 years, a little over 10 years. And when I was hired, the first responsibility I had was to be the general manager for then very nascent service called AWS Storage Gateway, which enables customers to bring virtually unlimited storage capability to their on prem environments.

So just as a personal note, this is an area that I've had a great deal of passion for, frankly loved for and understood its importance to you, our customers. And that importance is only growing as we, as we hear each and every day.

Now, we've also been building capabilities beyond AWS Storage Gateway over the years, for example, VMware Cloud which allows customers, is helping customers rapidly build their hybrid cloud operations.

In the last few years, we've continued building hybrid edge capabilities for customers who want to bring the same experience and benefits of AWS infrastructure services and APIs to wherever they want to deploy and operate their solutions, whether it's the traditional cloud, metro areas or edge locations.

And importantly, they want to be able to do this in a consistent fashion for all cohorts they support - for their innovators, for their builders, for their operators and their buyers.

And customers have asked us to go a lot further. As of today, we hear that they are facing challenges building one set of solutions that they can deploy and operate across the full continuum of hybrid edge environments.

And without a common experience across cloud, on premise and edge environments, it's slow to build and operate and it's complex and costly. And this is often due to a discontinuity in capabilities and experience.

Now everybody wants to overcome these challenges. That's clear from conversations we've had prior to this week and it's become even more clear in the conversations that I've had the privilege to be part of, maybe hopefully with some of you.

So why is that? Well, people want to be able to bring the business, they have the solutions, they have all the way to the edge, it's critical for their business. And while bringing the innovation that you all understand and experience in the cloud to wherever you want and enable location flexibility as you need.

Now, there are many applications we see customers using hybrid edge deployments for, but they generally fall into one of four major categories.

The first is due to latency. They need to make sure that whatever they're running has the lowest possible latency. So they want to be able to capture information, process that and make real time decisions at the edge or on prem. And this is true, whether they are fully connected to an AWS region, for instance, like Amazon Prime or whether they are disconnected from an AWS region. And what we often refer to is denied, disrupted, intermittent or limited connectivity or DDIL. So I'll use that term a few times today. So I wanted to define it upfront.

And you know what an environment like this? The International Space Station.

Next is local data processing. Customers want to be able to easily run data intensive workloads near where the data is located and processing results are used. And this helps them increase speed, efficiency and helps them frankly manage costs.

For example, we see a lot of customers that have a tremendous volume of data that they generate on prem at the edge in order to understand what is going on there. But they want to be able to filter that data before they move it to the cloud for further value extraction or for preservation purposes.

Now, a third use case is data residency and this is a use case that's really important for Fanduel in meeting its compliance and data location requirements. And when Fanduel gets out here later with Priya, they're gonna go into depth on what that really means.

But ensuring data residency where it is created while appreciating consistent cloud native capabilities and experiences is really important to a growing set of customers.

And finally the ability to enable migration and modernization of your applications. Whether you operate in a sports stadium or a factory floor or a container ship, customers want to extend or migrate their workloads to cloud in order to modernize them or bring greater value to the data that's created there by using services in the cloud as well as services on prem or at the edge.

So until recently, delivering the same experience from cloud to edge for innovators, for builders, operators and buyers and ensuring cloud benefits has largely been an unmet challenge.

And in this session, the primary purpose of this session today is we want to talk about how AWS is helping you overcome that challenge in practice right now.

And for the remainder of the session, we will demonstrate how we are helping customers by providing the same experience from cloud to on prem to the evolving edge regardless of where your application may need to reside. AWS is enabling customers to use the same infrastructure and services and tools to accomplish that. And we do that by providing a continuum of consistent cloud scale services that allow you to operate seamlessly across this range of environments.

Now, our objective is to enable your innovators, builders and operators and your business to focus on building your solutions once and deploying them to wherever you require it without compromise.

Okay. That's the beginning of the story. Now, what I'd love to do is introduce Jan Hofmeyer to start talking about how we're doing that today for you.

Alright, good morning, everyone. And thank you Werner for that first part there. Great to be here. Thank you for being here.

Tell me, have you seen Adam's keynote on Tuesday? If anybody here in the audience, ok great.

So, in Adam's keynote, he, he had a few great analogies of space and the vastness of space. And he spoke about those two explorer teams that went for a race to get to the center of Antarctica.

And you have to imagine that these folks must have an amazing passion for adventure and to get there first and to be able to do that. I kind of relate with those folks because I have a passion for mountain biking and I love adventure that most probably explain why when I went online and I didn't search for a comfortable mountain bike ride in Seattle. I searched for the toughest mountain bike ride, multi-day ride in the world. And I stumbled across this ride in Europe from Austria to Italy across the Alps for seven days that involved at least 10,000 ft of climbing every day.

And so clearly, the adventure part is there, but it was also those every day, every moment, every pedal stroke, there was always a challenge and you overcome that and you go to the next one and you get to the finish line and it's just an amazing feeling to look back after every day, after seven days and the accomplishments of doing that.

So if I look at, at the cloud, you and at AWS, we are absolutely passionate about bringing AWS and the cloud to where you need it. And it's not easy. There's a lot of challenges getting there.

You know, a great example, you know, in many cases, the challenges are easier to overcome. It's no different to when you migrate workloads from on prem to to a region and we can work with customers. But every now and then we, we hit really tough challenges.

A great example was Nasdaq. They were a pioneer in cloud. They started in 2008 migrating to the cloud and then they, they got to a point where they hit a really difficult workload and that was their exchanges. And how do you move that to the cloud? And the reason why that was difficult is in the exchange, there's a thing called a matching engine. This is for equal access trading. And the trading there is measured in microseconds, 20-30 microseconds.

So how do you support that in the cloud? So we worked with, with our customer there, partnered together, we solved that problem in Outpost that allowed them to continue their migration to the cloud, which was just fantastic to see.

So for the last few years, you know, we've been hard at work extending AWS infrastructure and services, new services to bring applications closer to where you need it, to where your users are, to where your equipment is, to where you process and make real time decisions at the edge.

And it starts with our regions, but we've gone further from our regions to major metropolitan areas, industrial centers and even locations within 5G networks. And moving on prem we offer solutions running in data centers in your offices, in retail locations. And we even going to extend it to remote and many times rugged environments and in these places many times with limited and no connectivity to the cloud and finally to devices, devices running in cities or automotive factories or wherever IoT devices are used today.

But let's start this all started with the region. You know, you've heard a lot about the importance of the regions and the cloud in Adam's keynote and many of the other keynotes this week here at, at, at, at Re:Invent. And AWS continues to be a leader in the cloud.

We have the most functionality with over 200 services, fully featured services in our region. Our continued pace of innovation. I'm sure you've seen here at Re:Invent in the last week, all the new services and features that we bring out and we are continuously investing and driving the innovation in the cloud.

And we are servicing the largest community of millions of active customers every month and more than 100,000 partners over 150 countries in what is today, the most secure environment that's available to supporting your mission critical applications and your sensitive workloads.

And of course, AWS has the most extensive global infrastructure. We have 30 regions with 96 availability zones and just to clarify what is a region, a region consists typically of three or more availability zones. Every one of those availability zones is a fully isolated and discrete data center that makes up the region and those are the regions that supports the more than 200 services.

And we've announced plans for seven more of these regions in Australia, Canada, Israel, New Zealand and Thailand. And this year, I'm excited that we've recently announced the general availability of, of new regions in United Arab Emirates, Switzerland, Spain, and India. In India this is our second region in Hyderabad.

But we've gone further and to bring our services even closer to where our customers need it. So let's take a look at how we brought our experiences and our services to customers that require data residency and locations where latency is really important for our customers.

When users access their applications, they're not only looking or expecting a fast and secure connection, but they customize and engaging user experience as well. And to help meet these expectations, AWS has built a cloud native content delivery network.

Amazon CloudFront is a highly distributed and scalable network that delivers static websites, apps, video, dynamic content like Amazon Prime with consistently low latency.

Amazon CloudFront today is available in 440 points of presence, 13 regional edge caches around 90 metros in 48 countries.

A great example of this was with HBO Max, one of our customers, they were looking for a content delivery network that could serve their ever expanding global customer base to deliver high performance and low latency. And they turned to Amazon CloudFront for its increased capacity and global footprint and edge locations.

And Amazon CloudFront very quickly became the top performing content delivery network for HBO Max. This allowed HBO Max to deliver the best possible video performance latency to their customers to enjoy their content.

Next up is Local Zones. So many of our customers are looking to do more than just content delivery. They're looking to run compute and storage solutions with low latency applications and local data processing. And for this, the Local Zones help our customers achieve this.

Today we have Local Zones. We brought Local Zones with compute, storage, databases and other select AWS services to large populations. And we have launched 25 Local Zones and have announced an additional 27 international Local Zones will become available.

And I'm excited to say that four new international Local Zones became available just last week. Now customers in Buenos Aires, Copenhagen, Helsinki and Muscat can enjoy or run their low latency applications in these new locations, supporting hybrid cloud migrations and the ability to meet stringent local data residency requirements based on their local regulations.

We support a number of services in our Local Zones today from launching your very familiar EC2 instances to attaching the same easy to use high performance EBS block storage to your instance to support your mission critical applications to running your container services with Amazon ECS or EKS and several networking services including AWS Direct Connect that allows you to establish a private dedicated AWS connection to your data center, office or colo location.

Environment. Mindbody is a great example of an enterprise migration use case where location and locality was important. Mindbody is a software as a service company that provides online scheduling and business application management software for the wellness industry. Like many other enterprises, they had workloads running in their existing on data centers with very strict um latency requirements between these applications.

AWS Local Zones provided the low latency between applications running in the in the local zone and their on premise data center using AWS Direct Connect in the local zone. And this allowed Mindbody to migrate applications incrementally and greatly simplify their migration to the cloud.

But we can, but we haven't stopped there at providing infrastructure at local zones and local metros. We also brought it to AWS to the telco five g networks to better serve the segment. We introduced AWS Wavelength to enable AWS infrastructure and services to run within a five g network wavelengths is available in 29 locations across the world. And we've launched in North America with, with Verizon and Bell Canada in Europe, with Vodafone in Japan, with KDDI and in South Korea with SK Telecom.

Here's a great example of a customer in Japan, with KDDI Makino Milling. Um Akina milling is a global manufacturer of milling machines in Japan. And they have these amazing autonomous mobile robots that moves around the factory, they move parts around and they place parts into machines and many times they would, these ams will collaborate with each other and hand over these, these parts as they go from machine to machine. And this is an environment where people are also operating leveraging five g network and the ultra low latency of the edge computing of a wavelength zone. These a rs are able to detect obstacles and move around completely autonomously. And it's just awesome to see, great to see how these new use cases become available and come to life with the ultra latency of five g with our wavelengths capabilities.

So as with metros and with telco networks, we have enabled the hybrid and edge applications to be deployed and operate within your on premise um helping you to run your migrations and modernization your low latency and your daily residency within your own location.

AWS Outpost is a family of managed solutions delivering AWS infrastructure and services to virtually any location on payment location. Customers can now run AWS native services on premise and it's available in a one year and a two year server and in a 42 year rack. And we can also have multiple of these 42 year racks to run a full data center with AWS Outpost.

One of the really important things and if I can leave you with one big message with outpost bu is not a server. How was is not a wreck help is AWS cloud running on premise. When you order an outpost, we deliver the outpost to your location. AWS will come on site, we will install the outpost, we're connected to the network that you provide that connects the outpost to to to the AWS region. And at that point, we will provision, provision it and hand it over to you as a customer. You log into the AWS console, you have immediate access to the same compute, easy to compute storage and network capabilities offered by outpost. No different to how you would use it in the region. So it's truly the region cloud. The AWS cloud experience delivered in your in your data center.

We've also delivered with the outpost server. Uh we had a new challenge with the outpost racks. You typically see customers with several data centers up to hundreds of data centers for for the outpost rack with the outpost server, you can truly take one a single server or multiple servers and install it in many more locations. And the intent here was think about retail, think about where you have thousands and tens of thousands of locations.

So it's really important for us to allow you to self install this server. So with a server, we we enable that where you order the server, we ship it out to the location, you connect it to the, to the internet and we will provision it and will show up in your aws console as easy to compute and storage on premise.

Its outpost is available in 73 countries and territories and as with local zones and wavelengths, many of the same services and services are available on outpost. You can run your favorite aws two instance on the outpost. We also support local on the outpost s3 for object storage. We have eks and ecs for container workloads, rds for outpost and elastic cash for highly available and highly performance databases. And we are constantly innovating and improving adding new features and capabilities. We've launched over 50 new critical features just this year in support of our customers. While we continue to focus on static stability, which is really important for us.

We recently announced the availability of local cluster for Amazon EKS on Outpost, a fully managed coin solutions from a so local clusters. Why that's important is to enable customers to now run the entire amazon cluster locally on the outpost. So customers can mitigate the risk of potential application downtime that could be caused by losing the connectivity between the outpost and the region. Typically happen when you have a fiber cut or a weather event and you lose that link, you can now run the entire es cluster locally. Um and that gives you the ability to, to for the application to remain available and the customer can perform the cluster operations during those network disconnects.

Now for a customer use case, I'm really excited to welcome Peter DeSantis, Senior Vice President of Technology at FanDuel onto the stage with my colleague, Priyanka Yanal, a Principal Technical Product Manager at AWS for a fireside chat.

Welcome, Shane. Thank you, Priya, great to be here. Tell us a bit about FanDuel and the business context that you operate in.

So FanDuel today is currently the number one sports betting company in the United States. We also have a daily fantasy business, a horse racing business, a gaming business. And more recently, we've just launched our new Fan TV proposition to bring the entertainment experience more to our customers.

F has been very fortunate in that we've been born in the cloud since practically day one of the organization and each one of those platforms has very different complex problems that it has to deal with. You know, think of a major e commerce platform. We have all the major hallmarks of having to deal with the peaky nature of our business. You know what is black friday and cyber monday for amazon.com, we have with the super bowl where we have to deal with 20 to 30 times. In some cases, the types of workload that we would get versus an ordinary day.

We also are affected like a bank, lots of money flows through our systems highly regulated and lots of complex problems. Generally when you deal with, with money, and we also are like a trading floor, we take our pricing and our products. We have to make sure that those prices and products are being calculated in real time based on sporting events from all over the world. That's a really complex set of problems.

We also have a really core focus of delivering responsible gaming to our industry. Um and, and that is, you know, part of our DNA, we try to build that in um across all of those different products. And when we deal with the scaling problems of all of those items from about 2018 onwards. And when sports betting was being legalized in the United States, you know, most companies were trying to figure out how to go from on premise data centers to the cloud.

We had the opposite problem. We actually needed to get out of the cloud and meet our regulatory and legal requirements where we were literally building data centers for the last 18 to 24 months in every single state that we operate in. That's not an easy task to deal with. And if you bring that back to the scale problems, the latency problems, the data residency problems, that's what makes FanDuel and AWS partnership really, really exciting over the last couple of years.

So given your need to run edge compute workloads, you've been an early customer of AWS Outpost service and now Local Zones. What were your top of mind business requirements? As you know, you went about evaluating these services.

Like most organizations agility and time to market is usually way up there as a business requirement, we need, we want to launch in a new location and at the earliest opportune moment and actually internally we have metrics, you know, the earliest allowable second. So if a regulator tells us that we can launch at midnight, we want to aspire to launch one second later. That's not an easy thing to do. But we want to set that bar very, very high for ourselves because, you know, early access into a market is critical for a brand new industry like sports betting regulatory and compliance is always top of mind. We want to fully abide by our regulatory and compliance needs.

And as I said, we want to be an industry leader in responsible gaming, making sure that we're doing the best that we could possibly do when when rolling out our products and services to our customers and capacity management and availability. Always a key challenge for most organizations building private data centers. And we wanted to provide the best customer experience, the fastest, the most reliable experience. And we wanted our customers to be able to use our products when they wanted to do it, not when it suits us. And that meant that we really had to think about how we were going to deal with the spike, the spiky nature of our business in the context of running our workloads.

And finally, I think performance and latency, you know that trading floor concept latency is really, really important from a risk perspective. If it takes more than a second, it takes too long. And when something happens in a sporting event, we need to know about it, we need to calculate recalculate and then distribute all that content down to every single customer that we have all across the United States, all in that highly distributed operating environment.

Um and I think, you know, when we look at all of that combined, there's many other requirements that I'm glossing over, but these are the primary four.

So what distinct benefits did um AWS Outpost and Local Zones helped you achieve.

So if we take those, those big ticket problems, faster time to market, that's a, that's the core, right? When we started to build our platform, it was taking about 12 to 16 weeks you know, think about ordering a data center, getting some hardware, installing that hardware, then installing your software testing it, getting it ready to go live and then flicking that switch to go live and without posts, we really challenge ourselves to take that problem and flip it on its head. How do we make it so that the installation of the hardware is the last thing that we do. And the seamless ability to work across region and in an outpost or a local zone means that we can pre deploy all our software, we can test performance test, really get the platform ready to go. So that the final piece of the puzzle is just redeploying our software. And we have some very creative engineers in FanDuel fantastic team who take these challenges, day in day out and they turn them into reality.

Great example, last year on the 31st of December, 2021 new year's eve, we had an AWS a team of AWS engineers go into our data center that was brand new in the state of New York. Eight days later, we went live to our customers that would be unheard of in the traditional on premise environment. And you know, when you look at what happened in those eight days, a lot of work happened before that. But actually, you know, getting the software to be deployed and do the final testing configuration was, was, was the the final piece of that puzzle and we have recently begun to deploy into local zones. And that takes that strategy, which is local zone is very complementary to that strategy, having the single control plane, having the hardware and the platform capabilities. But also wrapping that up into the data center. And the network means that we can go even faster. And we're going to constantly try to challenge ourselves to, to take that eight day period and shrink it down even further.

And we look at capacity management. Um when we were building this first, before Outposts was generally available, we worked with many hardware vendors to try and figure out, you know, how would we build this ourselves, you know, piecing together best of breed network hardware compute. Um but also trying to take a, you know, a build once deploy anywhere type mentality, cookie cutter, the entire environment that's complicated and requires an awful lot of investment in resource to figure out even simple things like getting device drivers working and getting the ecosystem correct so that all the software works, you know nicely with each other.

Um and throughout COVID, many of you, if you were buying hardware, you know, supply constraints was a big problem for us. Um just getting access to hardware by using Outposts. Amazon is, you know, using the same hardware in a region as they are with. Um you know, the Outposts in the Local Zones environment, we get to leverage all of the investment they've made over many years as well as having the logistical problem really, really solved for us. The AWS team really helps to, to focus on that delivery and we get to focus on building fantastic products for our customers.

Um I think Local Zones with even less hardware management for us, Jan referenced. It's not a server or a rack, that's what you order. But you do get Amazon really in your data center, it's a fantastic platform. And I think finally, you know, enhanced player experience with reduced latency, the more we utilize Outposts, the more we realize actually that we can create even more efficiencies within our application environment. Take that the approach of being able to put some services in a region, in some services in a state. You know, we've seen in certain examples where our applications will perform 30 to 40% faster by simply bringing them closer and running them on an Outpost and then architecting the applications correctly. So that when we're transferring the data back up to the region, it just works really, really well. And again, getting a 30 to 4030 to 44% performance improvement, you know, based on traditional off the shelf hardware. Again, that's a really really important part to that latency story

FanDuel is such a great example of how the AWS partnership and our use of edge computer services are helping customers like you scale. Um Shane, what does the future hold for FanDuel.

Yeah. Great, great question. Um I think it would be really important for me to just say we want to really continue the great work that we're doing that, the foundations that we've built over the last two years have really been instrumental in the success of us being able to launch new platforms. And we do want to extend it, you know, um the ability to use Local Zones as well as Outposts where available in any given state. You know, we deal with capacity management issues

"But actually how do we make that become a more efficient and sustainable model where we can take her to recovery sites, which is utilizing the same hardware, you know, how do we re architect and take the learnings from what is a region in aws and build our own miniature availability zones by combining the architectures together and that multi datacenter topology, i think is going to not only give more availability, more opportunities and provide that better customer experience and being able to deal with the peak and the bursty nature of of super bowl, you know, being able to just burst into a secondary data center is going to be really, really important as we continue to grow, you know, 5 to 10 x hopefully year over year.

Um i think lastly, the environment that we operate in, we're in about 17 or 18 states, it changes very regularly as we're launching and that means we have about 30 data centers or we will have about 30 to 40 data centers over the course of the next 12 to 18 months as we build out disaster recovery sites and those primary hosting locations that's going to put an immense strain on our network because that is how we connect all those environments together is the next set of problems that we need to solve for. It works today. But, but how do we get it so that it's multiple gigabits of transit and being able to securely manage all those sites. And we're actively looking at things like aws cloud one as a way to really simplify the management and the overhead of that network and really leverage the rest of the aws ecosystem. So some really exciting things coming.

It is so great to hear how you have been able to reduce your time to market, reduce costs, navigate regulatory requirements all to reimagine your business operation. Thank you so much for sharing your story, shane. And thank you very much to everybody for listening. I wanna thank uh shane and priyanka for that amazing use case from fanduel demonstrating how the power of the continuum from cloud to edge is working for them and their customers. And for those of you who love to support uh bet sports uh for some of you as well.

So so far, we've looked at how we can bring the same value and experience to regions like for companies like hbo max and to metro areas and telco organizations for like mind body and on premise data centers like fanduel. Uh now we're going to look at how we do this for the remote rugged and limited connectivity locations. So many of you have hybrid edge needs that need to be deployed to solutions that are mobile or intermittently connected or even disconnected environments. So how do we continue, how do we continue to bring this value uh of hybrid edge all the way out to the ever evolving edge. We do so with the snow family.

Uh the snow family of devices serves customers who need to operate in extreme rugged, denied, disrupted or limited connectivity environments, for example, construction sites or oil platforms or mobile environments such as ships or cars or airplanes or rapidly deployed military forces. And uh we even recently deployed aws infrastructure and customer applications to one of the most extreme environments to date, which is the international space station. Now the snow set of devices is available in 20 of our 30 public regions and we're constantly uh launching the devices in more and more regions. So you can expect to see them uh continue to to to be launched in in more regions as we go go forward. And this enables you to deploy aws based solutions to capture process and make real time decisions and move large volumes of data just about anywhere on the globe to an aws region for further value creation and retention.

Now within the snow family of, of product, uh i'm going to talk about two of our, our family members that are specific for edge computing. And that's the aws snow cone, which is about roughly about a £5 little less than £5 device about the size of a tissue box and aws snowball edge, which is uh osha compliant less than £50 device about the size of a carry on suitcase. And each member of this family can operate in extreme conditions and have passed a, an array of military and dod uh uh standards for data durability, uh data security, physical durability and physi physical security.

Now, these devices can be operated in stand alone mode or in the case of snowball edge, they can be racked and clustered for high availability applications or when you want to extend or make, have more capacity for compute storage and networking. And each is a self contained unit that can be shipped anywhere in the world. It it contains an e ink label that allows that to happen. And it's really, really simple to order and receive. You simply go to the aws console, you select one or more devices that you would like, you select the application that you've developed and we will deliver that device fully ready to go when it lands in the region or location that you want.

And, and in the snow devices all snow devices can operate in these denied disconnected, limited or inter intermittent connectivity environments. Now as with aws local zones, wavelength outpost, many of the same services for building in those environments can be done. So on snow family as well, enabling you to build, deploy and operate with a consistent set of capabilities and experiences. And this enable enables you to operate wherever you'd like to operate on ad dil uh edge environment.

So for example, you can build applications using ec2 virtual machines or ek ss containers or lambda functions using ebs storage or s3 storage. You can use them to uh take machine learning models that you've trained in the region and deploy them to the edge using sage maker edge manager. And with that you can perform such activities such as image processing at the edge. In fact, this is exactly the application that runs on the international space station or you can collect and process iot sensor data with iot street street wise, sitewise, excuse me, edge providing real time factory floor equipment monitoring now.

And based on your feedback, we continue to innovate and iterate on the services that are that can be deployed with the snow family of products as well as how you can manage and administer your applications and your environments on the edge.

Now, for a customer example, um we're very proud of what we enable uh for customers along the entire continuum, but there are specific applications where customers need to be able to operate environments both in the cloud and on the edge because of discontinuity and, and communications and, and, and uh help ngo is a good example of this. They are an international organization specializing in emergency response preparedness, risk mitigation and prevention and their forte is improving disaster response results through technology and central to help ngo s uh technology.

Um you know, kit if you would is mobile edge capabilities using snowball edge. Now, as example, in 2021 a tornado hit kentucky and destroyed the local communication systems completely and that prevented help ngo from analyzing drone imagery of the of the disaster area using an aws cloud based application. They already had running to do so. Now using snowball edge help, ngo was able to quickly deploy that application as it developed in the region onto a snowball edge, which allowed them to do local processing of that drone imagery right on site.

They were able to take a cluster of snowball edges, deploy them to the environment and run parallel processing on 20 hd streams that came off of those drones. And what do they do with that? It allows them to target responses to where it's needed most because prior to this type of technology, you would send people out to the disaster relief area and try to figure out who needs help. Where with drones they were able to do that automatically with image processing using the snowball edge cluster, they were able to do it in real time on site without the need for the standard connectivity environment, they would have uh from cellular wi fi and other sources.

So this story directly demonstrates how aws enables a consistent set of capabilities and experiences from the traditional cloud all the way out to the remote disconnected edge. And if you guys have a few moments, uh i know it's a busy show. But if you have a few moments today, head over to caesar's palace and go to the forum uh where we have a tent there and you will actually see one of these jeeps and these are amazing uh uh uh pieces of technology because you know, these are packed with uh edge devices, coms gear, iot sensors, a snowball edge uh and sat com gear and software defined communications gear which enables them to join public five g networks or to stand up a private five g network when no communications uh exist in the area that they operate. So if you have a moment, you can run over there and you can see the jeep. It's uh it's a lot of fun.

Ok. So now let's talk about extending cloud service. It's the smallest edge with smart devices. There are more than 14 billion smart devices in the world today. And it's often in things we think about like wrist watches and cameras and you know, cell phones and speakers for that matter. Um but more often it's the stuff you don't see every day powering industries of all types and for all types of customers, for instance, nascar uses smart devices to provide real time data for driver insights for race strategy and analytics, which is pretty cool. And hill corp which is a leading energy producer uses smart devices to monitor the health of their wells to optimize production and to proactively uh predict failures which allows them to improve on their capex spend.

Now, aws iot provides a deep and broad set of services and partner solutions to make it really simple to build secure managed and scalable iot applications. And we and, and all we've talked about today for our iot based applications can be built deployed and operated on outposts or snowballs and can also be operated on a whole host of third party gear.

Now we help customers choose the approach they want to take based on their technical capabilities, their customization requirements and their speed to market needs. And uh one of the spectrum customers and partners can build on our foundational services and you know, a lot of our foundational services be a compute networking storage, but as well as ml analytics and iot foundational services such as green grass and green grass is an open source runtime environment that allows you to build these services and deploy and manage device software at the edge.

At the other end of the spectrum, customers can use ready to deploy solutions that are built by both aws and our partners to quickly realize business value. And in the middle of the spectrum is purpose-built iot services designed to speed time the market in very specific industries, uh and use cases such as robotic orchestration and industrial monitoring.

Now, one such service is iot sitewise which is specifically built to operate the edge, helping customers track and monitor industrial equipment at scale. And we continue to build purpose-built services for you including iot twinmaker which helps create digital replicas of real world environments. Uh iot fleetwise, which enables you to collect tremendous amounts of automotive data and bring that to the cloud for value extraction and aws iot roborunner which enables fleets of robots to operate together at the edge.

Now, an exciting example uh comes to us from john deere and john deere uses iot in two ways. Uh the first is for for precision agriculture and the second is for improving factory operations. Now, in the first case, uh john deere uses iot analytics technology to track report and manage a fleet of over a half a million uh pieces of equipment and they help farmers get the most out of their harvest by doing so. Information that is collected using aws iot is displayed on dashboards that john deere maintains on mobile applications that farmers carry around and on the monitors within this half a million uh uh uh half a million uh devices or uh pieces of equipment uh that the solution is deployed on. And farmers use this information to generate more food per piece of land while simultaneously reducing the amount of pesticides, herbicides and consumables they use to produce our food.

Now, the second use case is john deere is using green grass, sitewise twin maker and amazon uh chime sdk for connecting factory operators with experts all around the world that help john deere produce these great solutions. They use green grass in sitewise to enable them to monitor data. They use twinmaker to provide operators the ability to show experts exactly what they are seeing and they use chime to enable a real time dialogue between these operators and these experts.

So with this solution, now operators can request assistance from experts located anywhere in the world. Those experts can quickly uh using this uh multiple data sources including the monitored data sources from sitewise, the data that they're getting from uh digital twins and the data they're getting from the conversation they're having with the operators to troubleshoot what the operator is seeing and be able to quickly generate the results of the operator needs. And you know, for me, this is a super exciting uh solution because they're combining multiple foundational and purpose-built services along with multiple data streams and amazon chime into a whole solution.

Ok. Now that we've seen how aws is delivering a consistent experience from cloud to the furthest and smallest edge i i want to take a moment to look at an example from a customer that brings them together bringing multiple elements from the cloud edge continuum to their solution.

Coach base is a non sql database company that uses aws hybrid edge services such as local zone wavelength and outpost and snow family to deploy their applications in a highly scalable, reliable and performing environments to reduce latency by over 18% for their customers.

Couch space is a no sql database company and we build databases for highly scalable, highly available applications that span the cloud to the edge. When you're trying to build highly available applications that are highly performant. What you really need is you need compute networking and a database built for highly distributed topologies. And the great thing about what's going on at aws is the optionality they provide for that compute tier. And so what we built on this, these aws compute environments was highly distributed managed or self managed database for the cloud, an internet gateway for accessing that data securely over the web and synchronizing that data down to the edge and that works across the cloud, the edge and on the offline first compute environments. And if you look across the rest of the industry, that type of optionality just doesn't exist.

And today our customers are using aws local zones, outposts and wavelength and with aws local zones, the data center and the hardware is managed by aws, aws will manage the life cycle of those systems. And i can just take advantage of the services, the speed, the latency and the availability that i need for my applications. And the great thing about aws outpost is you can put this anywhere and with aws wavelength that really unlocks the power of what you can do with five g. And now developers can just log into their accounts, spin up their database, spin up their compute across the world and then just focus on their applications building amazing high available applications, they're always fast. And when my teams did engineering tests, moving our compute and data down to the edge from the region, we saw 80% reductions in the latency when the system was accessed from the edge versus accessing it in the region. And so when our customers are able to build applications that span the cloud to the edge, they can focus on building their applications iterating over time and delivering on the needs of the business. Excellent.

Well, uh let me take us home here. Um great to uh first of all, i want to just thank uh wayne um and uh peter desantis from fanduel, a great story. Great use case. Uh and then my colleague uh priyanka for uh for their sessions. I'm going to quickly cover one more use case if you give me a few more minutes here of that just shows you and the important thing here is where you don't select one location. It's not that you have to pick, i'm going to run in the region. I'm, it's really picking the right location for the right workload. And, and five g is, is, is a great example of that. If you look across your public, your private and you make, which is a different parts of the, of the five g network.

So the first one, let's look at public five g. Um you know, after dish announced the launch of their service built on on, on aws, they leverage our regions where they deployed the five g control plane which is not latency sensitive, that's used to control the network. And they deployed that in our regions and in the local zones is where they deployed the five g user plane, which is the data plane itself, highly sensitive to latency needs to be closer to the towers. So that's a great example of where they leverage the regions and the local zones.

Similarly, swisscom in europe have selected aws to build their next generation five g network leveraging the new zurich region in switzerland and working very closely with aws as they develop a cost effective highly available solution within, within the region.

Going to the private. We have several solutions today running on awe as providing private five g, one of them, which is the aws private five g that we announced last year is another great example. We wanted to make it really easy for customers to be able to deploy, operate and run their private five g or private wireless networks and using the cloud economic model where you pay as use capacity versus paying for devices. So that is a great example of, of us using running the aws 55 g in our regions, we run it in our local zones and we run it local within the outpost.

We are also working with telcos and with technology partners for them to develop their private five g solutions running on aws. Um this is to enable them to run closer to the edge and make this available a great example is park new zealand um where they used actually the snowball edge that allowed them to create a highly portable edge solution, they can literally fit in a suitcase and uh they can ship that out and then stand up private private wireless networks in remote locations.

The next one is the multi access edge computing or they call it mac for short. And again, make as you can imagine how you can bring compute in the network in the five g network closer to, to the user. You have the wavelength zones that we cover today that brings it to put it in the network, you can take it with outpost and move it on premise. So make us another example that uses those two components to get closer to the user.

So all of this shows right, how aws is today the best place to run five g networks and not only five g networks, but many other workloads. As you heard from fanduel and many of our other customers aws help our customers innovate, build, deploy and operate in the hybrid cloud and the hybrid edge in a consistent with the same a ps the same tools across the entire continuum of the cloud we shared with you, how we accelerate innovation with the broadest and the deepest set of cloud to edge capabilities, how we create a truly consistent experience across the environment for builders and our operators and along the entire cloud edge continuum, how aws can help you optimize your agility and results and experience with the best partner support.

And i have to at least finish here at saying as we have an amazing set of customers and partners and i want to thank all of our customers across the hybrid cloud and the edge computing for your trust in partnering with us as you develop innovative solutions for your customers.

So with that, thank you everyone."