 Quadrant for the infrastructure as a service segment, and you can see that AWS is in that top right corner, and the other 14 companies are clustered somewhere in the middle, and their estimate is that AWS has more than five times the compute capacity deployed than the other 14 providers combined. So we have a significant leadership position, and we're growing really rapidly. We're going through a shift in technology that is unlike any other in our lifetime, and it's happening at a startling pace, way faster than anybody anticipated. And you can see a few quotes here, here are some Gartner quotes, enterprise customers are skipping the years of early getting their feet wet and immediately jumping in with more significant projects with more ambitious goals, or increasingly organizations are asking what can't go to the cloud rather than what can. Or as 2014 dawns, we're moving into an era of truly mainstream adoption of cloud. So the question that lots of people ask us is why are people moving so fast to the cloud and to AWS? What's driving it? And we think there are a few factors that will explain in a little bit more detail. Probably the number one factor in companies moving to the cloud now is the agility they get. So why does agility matter? What's the big deal about it? Well, if you look at some of the industries that have been completely disrupted by brand new companies who've largely built on top of the cloud over the last few years, you get a pretty good idea. You know, look at what Airbnb has done in the hotel space, or Spotify has done in the music space, or Dropbox in the consumer storage space, or Instagram in the photo space. Those were long standing industries with long standing leaders that were completely disrupted by these companies that started from a standing start. And so what it means is that enterprises and businesses can't afford to be slow. Yet, if you go and you talk to an engineering manager, an enterprise, or an engineer, and you ask, how long does it take for you to get a new server to try and experiment or to put something in intake? The answers we get when we ask that question range from six to 25 weeks, and typically they're 10 to 18 weeks. So if you think about that, it's maddening. If you're an engineer, you're not going to bother spending time trying to invent because it takes so darn long to get a server. And God forbid you're experiment successful, how are you going to get all those other servers to actually roll it out? And anybody that does a lot of invention will tell you that the two most important things are, number one, you have to be able to try a lot of experiments, and number two, if those experiments fail, and many of them should fail if you're really pushing the envelope, you don't want to have to live with the collateral damage of failed experiments. In the cloud, that world is radically different. You can spin up thousands of instances or servers in minutes, and then if your experiments don't work, you can either give them back to us and stop paying for them, or you can reuse them for other experiments. It's a very different model than what's existed the last several decades. And what it means is it allows you to move into a different world. So now that you're in a world and a reality where your employees know that there's a possibility they may be able to deploy more quickly using the cloud, all of a sudden you get a movement from learned helplessness to actually getting some shower cycles from your employees as they think about how to improve the customer experience. And it also allows you to take that innovation and that group of people thinking of new ideas from just a small group, typically at the senior levels who have some visibility into how they might ever be able to deploy, and with the cloud, all of a sudden everybody is incented all over your organization to think of new ways to improve your customer experience. As we have been building this business over the last several years, one of the questions that we get asked a lot is what happens to my IT team with the cloud? And of course the IT team doesn't go away, it just works on more value added activities on top of the infrastructure instead of the undifferentiated heavy lifting of this racking and stacking. But a question that we don't get asked as much, which I think is an even more important question, is how do we as businesses empower and incent more of our employees all over the company to be inventing on behalf of our customers? The cloud unleashes all of that innovation. It's great for customers, it's great for companies, it's great for businesses, it's a win across the board. So the second reason that we see people moving to the cloud so quickly is the very broad functionality that exists in AWS. And so whether you're building an altogether new business or new application or you want to find a way to move large swaths of applications or all of your applications over a couple year period to the cloud, you need very broad functionality to make that happen and to be able to do it in a coherent fashion. And that's what we've been building here with AWS over the last eight years. We have a very broad robust technology infrastructure platform that we've been iterating quickly on and that has a lot more functionality than you'll find anywhere else. So I'll just give you a little bit of detail about it. So at the lowest level of our platform, you have the infrastructure pieces. The way to think about regions in AWS is a region is a physical place in the world where we have multiple data centers or availability zones. So we have regions, we have availability zones, and then we have points of presence or pops for our content distribution network, cloud front. And then we have a number of building block infrastructure services in compute, storage, and database categories. So for compute, we have Amazon Elastic Compute Cloud or EC2, which most people know about. We have the ability to use load balancing with that called Elastic Load Balancers or ELB and to autoscale. So you can set triggers for when you automatically want to increase the number of instances or decrease the number of instances in your deployment. We have a number of storage services, including a very large object store called Amazon Simple Storage Service or S3. We have a block store called Elastic Block Store that goes with our EC2 computing service. We have a backup and archival service called Amazon Glacier, which we launched about a year and a half ago. That charges about a penny per gigabyte per month. And then we have a number of database services. We have a relational database service called Amazon RDS. That comes in flavors of MySQL and Postdress and Oracle and SQL Server. And then we have a very high throughput, very low latency and fast, no SQL database service called Amazon DynamoDB that we launched a couple of years ago. And then we have a cache and service called Elastic Cache. Then we have a number of networking capabilities on top of these services that you can use. So we have something called Amazon VPC or Virtual Private Cloud that effectively allows you to cordon off a portion of our network and deploy resources into them. And then you can set up subnets. And some subnets can be addressed exclusively through IP addresses that you bring to bear. Some subnets can be set up to be accessible through the internet. And then it connects to your own data center topology through VPN. So most of our enterprise customers are using VPC. And then we have something called Direct Connect, which is a private connection between your data centers and AWS. And then we have a DNS service called Route 53. Then we have a whole number of analytics services that we've been building over the last few years as more and more customers have told us that they want to use the cloud and the advantages of the cloud, how easy and inexpensive it is relatively to do more and more analytics. So we have something called Elastic MapReduce or EMR, which is really Hadoop on top of EC2. We have something called Amazon Redshift, which is our data warehouse service we launched a year ago, which is the fastest growing service in the history of AWS. We have a new service that we announced and launched at Amazon Reinvent this past November called Amazon Kinesis, which is a new platform for real-time processing and streaming data. And then we have a number of application services. So we have a content distribution network called CloudFront. We have a notification service called SNS. We have a queuing service called SQS. We have an email service called SES. We have a workflow service SWF. We have a service that we announced at Reinvent called Amazon Workspaces, which are virtual desktops in the cloud. We have an application streaming service that we announced at Reinvent this past November called AppStream. And then we have a search service called CloudSearch, so a broad array of application services. And then we have a number of management and administration capabilities. So we have a monitoring service called CloudWatch. We have identity and access management that allow you to have fine-grained controls for the different users in your organization. And then whether you want to work at the API level, the command line level, or with a web GUI with our AWS Management Console, we have all of those options available for you. And then we have what people traditionally, I guess, have called platform as a service or pass, but we have a number of services that are really higher level abstractions to using our platform. So we have a container service called Elastic Beanstalk, which comes in a half dozen languages, which is really a drop-in container. You upload the file, it automatically spins up the instances, sets up the load balancing and the auto scaling and the monitoring. So very little work for you. Then we have something called OpsWorks, which is a DevOps tool that allows you to do pretty broad and expansive orchestration of your environments. And then we have something called CloudFormation, which are really like templates or recipes of stacks that you build and that are easy to both stamp out and then deploy. So a number of these services are this pass layer. And then we have a very broad ecosystem. I talked a little bit earlier about our systems integrators and ISV partners. We also have something called the AWS Marketplace that we launched about two years ago. This is a place where our ISV partners are able to offer their software already configured to run on EC2 so that you can actually easily deploy those either in one click or you can do it on your own at your own pace. This is a very fast-growing area for us as well. We have 25 product categories, over 1,400 product listings, and it's grown about 90% very quickly. And then we've built a number of people-oriented capabilities in the last few years as our customers have asked us to do so. So we have a premium support function that allows you to have one-on-one technical account managers. We have professional services where we have a team that comes into your company and helps you think through how to move to the cloud and which applications are best suited to move in what order. And then we train tens of thousands of customers all over the world and also provide certification exams and badges for them. So a very broad, very robust technology infrastructure platform that's continuing to get broader quickly and has a lot more functionality than you'll find anywhere else. But it's not just having a lot more services that helps you. That is obviously a big help. But it's also what number of features and capabilities you have within those services because you can launch something that's very feature poor and it doesn't allow you to accomplish your mission. So you typically need a lot of capabilities in these services. And we also are iterating and adding features at a very fast clip. And so this is an example just with Redshift or data warehouse service. So we launched that a year ago or so last February and already just since last February in 13 months there are 56 new features alone in Amazon Redshift. So iterating at a very fast clip. Here's another example of where the amount of features and capabilities within services matter. So we have a number of instance families in EC2, a very broad number of those. And we have those because we've been listening to what our customers tell us they want to optimize for their various workloads. So way back when we started with our standard instance family, which is the M1 family, which has a balance of compute and memory and network. And then we had customers who said, boy, I have compute-intensive applications. Can you build a compute-intensive instance family? And that's why we built the C1 and the CC2 families. And the only difference being CC2 or cluster. And then we had customers who said, I have IO-hungry applications. Can you build an IO-optimized instance type? And that's why we built the HI1. And similarly, we had customers, especially data warehousing customers and those running Hadoop who said, I have really storage-hungry workloads. Can you build a high-storage instance family? And that's why we built the HS1. And then we had a number of graphics companies and gaming companies and scientific computing companies who wanted a GPU in their instance. And so we built the CG1 instance family, which has GPU inside of it. And then we had customers who wanted memory-optimized instance families. And that's why we built the M2 family and the CR1 family. Again, the CR1 family differences that they're clustered. But over the last six to nine months, we've spent a lot of time building a new generation of instance families where we are able to provide the latest and greatest and the best possible performance for all these types of instance families. So for instance, the M1 was refreshed to the new generation M3s a few months ago. And the M3s have SSD-based storage and they use all the latest processors. And they have much better and more consistent performance at a lower price than the M1 family. The C3 family, similarly, is the latest instance family in the compute-intensive world. And this instance type has enhanced networking. It allows you to have four times better packets per second than the C1 instance. It has hardly any jitter, so the performance is very consistent. It has also about 60% better overall performance than the C1. So this instance family has been very popular and people have used it to a very large extent. I'll also add that what's interesting, too, about the C3 instance family is it not only has this great enhanced networking, but it also has a significant amount of memory. And it turns out that our customers who are running high-performance applications often want and need that memory to have better performance. So this has been very, very popular. In the IO-intensive world, we've added a new family called the I2 instance, which is an upgrade from the H1 generation. And this also has SSD-based storage. All of our new instance generations have SSD-based storage. And it has more SSD-based storage than the prior version, about 56% more than the H1. And it has much better IOPS capabilities. So on the largest instances of I2, you can get 365,000 4-kilobyte random read IOPS and 315,000 4-kilobyte random write IOPS, so very strong IOPS capabilities. We launched a new G2 family that has more instance sizes than the old CG1 at lower prices and better performance. That's a win for everybody. And then we have two new announcements we're making today of things that are coming in the next few weeks. So the first is a refresh of the high-storage instance family, the HS1s. And this refresh will have double the VCPU threads, and then it'll also take the storage throughput from 2.6 gigabits per second to 3.6 gigabits per sec. And then we're also going to announce, we're also announcing we're going to launch a new storage optimized family called the R3, I'm sorry, memory optimized family called the R3. And this will be the best place to run memory optimized workloads in AWS. And so this will have SSD-based storage, like all the new instance families. It will have an 8 to 1 ratio of memory to CPU. You'll be able to get up to 244 gigabytes of RAM, and it'll have about a third better performance than the M2s did. So the R3s will be significantly better than the M2s in performance, and then it has all the same great enhanced networking on the R3 side that the C3s have and the people like so much. And these are coming in the coming weeks. And so we've taken all these instance families and then we've connected them to our world-class network. So when we started AWS about eight years ago, we had a very strong network because we needed that to grow Amazon, the very large retail business, the first 10 years or so we were in business. But as you can imagine, with the explosive growth of AWS and becoming the largest cloud computing platform in the world, we've had to substantially build out and hone our network over the last eight years. And we are relentless about working on our network and continuing to provide better and better performance. So with this world-class network, you get high packets per second. You get low jitter. You're able to get a dedicated connection, both network connection and bandwidth from EC2 instances to optimized EBS instances. You have various virtual network interfaces, very high throughput, low latency network. And then you're also able to designate having instances close together proximity-wise for even better inter-instance network latency. So having this type of platform breadth involves having a large number of services, a lot of features within those services, but it also involves having a very broad geographic footprint. Most companies these days have end users all over the world. And in fact, most startups launch with end users all over the world. And as such, companies don't want to have to settle for what was the case in the old days, which was it was too hard and too expensive and too unreliable to try and find infrastructure providers in other countries that were closer to your end users. So they pretty much served everything out of one location in the world. That doesn't have to exist anymore because of the very broad footprint of regions that AWS has. So we have 10 regions at this point. We have regions on the East Coast of the United States, on the West Coast of the United States, in Europe, in Singapore, in Tokyo, in Sydney, in Brazil, in China, and we also have a region called GovCloud, which is a government-only region for ITAR compliance workloads. So now in this world, when you think about having platform breadth, it means that you can actually deploy your application in any of our 10 regions in minutes and it allows you to have better latency and a better customer experience for end users all over the world. So when you look at this combination, when you can spin up thousands of servers or instances in minutes and you can draw from all of these services and all of these features within these services and be in all the regions that AWS provides, what it means is that you not only can get from having an idea to market or deployment in the shortest possible time, but it also means that you can move most and or all of your workloads to that cloud provider in a relatively short period and enjoy all the benefits you get in the cloud. So I thought I would give you a few examples of how some customers are using this functionality to get a sense of things that before were either not possible or really difficult to do that are now much simpler given the platform breadth AWS has. So let's start with Supercell, which is a very successful gaming company that has about 100 employees, but a very large number of users, eight and a half million daily users just in iOS alone, and they have several hit games from heyday to Clash of Clans. And if you think about what gaming companies care about, they are constantly trying to analyze and measure what their users are doing so they can make the user experience better and also so they can monetize better. And so Supercell with Clash of Clans is using AWS and what they do is they take all of the customer data, all the taps, all the clicks, everything customers are doing, and they stream those into Amazon Kinesis, which is our new platform for real-time processing of streaming data. And then once they have them in Kinesis, what they do is they have a whole set of processing applications built on top of EC2 that is analyzing that data and running various processing jobs on them. And then after doing so, they take that data and some of that data they bring to an internal dashboard or visualization at headquarters so that everybody at the company can visualize and see some of the hot trends that are happening in the game. But they also take all that data and they put it in our storage service S3 for further analysis. And then what they do is they use our data warehouse service, Amazon Redshift, to do all that deep querying and analysis they want to do on what their users are doing and how they want to feed that back into the game. And their business intelligence users can use the same business intelligence tools they've been using for a long time because all the BI tools are integrated with Redshift. And then for some of the data that they actually don't analyze regularly, they put that into our backup and archival storage service, Amazon Glacier, which stores data for one cent per gigabyte per month so that if they want to query on that data later or use it in whatever fashion they can imagine, they can do so. And because it's so inexpensive to store in Glacier, we find that companies like Supercell and lots of others are storing data that they regularly had to dispose of. So let's take another example. Let's take NASA. So NASA has been working with AWS for several years now and NASA's had this Mars rover. It's been taking pictures on Mars for a long time and they would take those pictures and send them down to AWS, store them in S3, use EC2 to do a bunch of processing and then send instructions back to the rover to move forward from there. But when they were actually having the Mars Curiosity land on Mars, not surprisingly, a lot of people wanted to watch it in real time. And so they decided to run and stream that live landing on top of AWS. And so what they did was they set up, they basically went to the AWS Marketplace and they found an offering that was Flash Media Server, FMS, on top of EC2. And they decided that's what they would use to do the streaming. And then they set up a number of those FMS servers on top of EC2 behind elastic load balancers to make sure that they could handle the load across different regions and AZs. And then they used Route 53, our DNS service, to do the routing to the different clusters of them. So they used weighted round robin distribution that allowed them to designate different load to the different clusters that had the most remaining capacity. And so each of these FMS servers they figured out could handle one gigabit per second. So they replicated them and they deployed them to multiple availability zones in multiple regions so that each cluster of FMS servers were able to handle 40 gigabits per second. And what they then did was that they used our service called CloudFormation, which is this very powerful automation and authoring tool to set up the stacks that they wanted to deploy so they could just stamp out these 40 gigabit per second clusters of FMS servers. And it deploys them in the order that you decide to orchestrate them and you can do it in a repeatable way so it was very simple for them to do. The whole thing was about 181 lines of JSON. It was very quick for them to write and then to test and deploy. And what they found was that as the volume became larger and larger, which is of course what happened as people tuned in to watch the landing, it was really straightforward for them to just stamp out and deploy these CloudFormation clusters. And of course the landing happened, their streaming went flawlessly and they were very pleased with the experience. Here's the third example. So Nintendo, the gaming company, has a lot of gaming companies with popular games like Mario Kart and Pokemon X and Y. They have a social network, social networking app really embedded inside of all of these games. And you need a few things if you're going to have a social networking app that's embedded in these games. And that social networking app, which is called Miiverse, is runs on top of AWS. And so to make that work right, you need three things. The first is you need very low latency everywhere in the world. And so what Nintendo has done is they've used Route 53 or DNS service and they've used latency-based routing capabilities. And what that does is it routes requests to the closest, fastest possible location. So it doesn't always take the region or the pop that's closest to where the request comes from because it also considers the internet traffic and what's happening across different connections during the day. So it takes the fastest possible location to route those requests to. The second thing you need, apart from low latency, is you typically want to do a lot of analysis on top of what's happening and what the social networking activity is within the game. And so you see that Nintendo uses Redshift, our data warehousing service, to do all that analysis. And then the third thing you need is you need very high availability. And so what they've done is they've deployed in multiple regions in multiple availability zones and then they've connected and they've deployed those in VPCs in virtual private clouds and then they've connected all those instances through a private virtual network and they're using a capability that we have in VPC called Elastic Network Interfaces. And what Elastic Network Interfaces allow you to do is decouple the instances themselves from the network configuration and topology, which means that if there's any kind of hardware or software failure, you're able to swap things in without missing a beat. So these are three very good different examples of where companies are using the very broad breath and capabilities of the AWS platform to do things that either were impossible before or would have taken so much effort and energy that doesn't differentiate their experience. And when you look at some of the services that these three examples are using, Amazon Kinesis, Amazon Redshift, Amazon Glacier, Route 53, weighted round robin distribution, latency-based distribution, cloud formation, AWS Marketplace, Elastic Network Interfaces, when you look at all those capabilities, they don't exist anywhere else. So having that very broad platform breath is very important to customers and allows them to move more quickly. So people are moving really quickly to the cloud because it provides them better agility and because there's all of this breath that not only provides them agility, but also lets them move all of their workloads in a coherent, quick fashion. But a third reason they're moving so quickly to the cloud is that you get continual evolution, innovation, and iteration in the cloud versus what you might do on-premise. I think if you talk to the AWS team and ask them collectively what they're most proud of other than the security and the operational performance the first eight years, it would be the pace of innovation. And if you look at, say, 2011, we launched what we considered 82 significant services and features. In 2012, that number was 159. In 2013, last year, that number was 280. And this year, year to date, we're ahead of last year's pace. So the pace of innovation at AWS is accelerating. But also, if you think about trying to manage an infrastructure on-premises, it's a lot of heavy lifting. You have to build out all of those services yourself. You have to hire all the resources and the right types of people who know how to do that. You have to manage it. And typically, it's expensive and it's slow. Now, if you compare that to what you get in the cloud, you get all of this for free. So if you look at the right side of that slide, you can see that this is just the features that launched in the week of March 3rd, 2014. All of these came without anybody having to build anything themselves. And it came overnight and people wake up in the mornings and they have new capabilities they can take advantage of. Some of which just automatically happen in the service and you just get the benefits of whatever we improve. Some of which are there for customers to decide how and when they want to deploy. But this is a very different world than having to maintain all this stuff yourself on-premise. And I'll give you another example of this that illustrates it. This is DynamoDB, which is our very fast, high throughput, low latency, no SQL service. And we launched this about two years ago. This is one of the fastest growing services we've ever put out. And right from the get-go, it had Internet scale throughput and very consistent performance on the iOS side. But not surprisingly, we added a lot of capabilities, as you guys told us, what mattered. So we added global and local secondary indexes to improve the query flexibility. And that was a really big deal for our customers. And then we added item level access control that allows you to put an access control policy for an item and a table. And then throughout the year, we add all of these other capabilities that, again, you just get for free to use however you want, ranging from parallel scan and batch rights and a geospatial indexing library and a number of testing tools that you can operate locally and availability in GovCloud and price decreases. All of this just happens and evolves for you when you're using a platform like AWS. You see it not only with iterating on existing services that we put out, but you also see AWS continuing to innovate on altogether new services that didn't exist before. So I'll give you a couple of examples. This is a service that we launched and reinvented this past November called AppStream. And the problem we were trying to solve was that with the explosion of mobile devices and mobile usage, a lot of developers were left with an uncomfortable and unhappy decision, which was if they had graphically intense applications or experiences, they either had to send it verbatim to the small number of people that had devices that could handle those graphically intense experiences or they had to water down the experience to get mass adoption. And our customers asked us if there was a way for us to try and solve that issue. And that's what we do with AppStream, which is an application streaming service that allows you to stream either entire applications or components of applications that are graphically intense. And then we've built our own protocol here that recognizes what the network connection is and what the device is and adjusts on the fly. And you can see these are three examples of customers running on top of AppStream. These are all running on top of AppStream, as we speak. And I'll give you a couple. Eve Online is a very popular massive multiplayer game. It has, you know, 500,000 simultaneous players. And one of the things they do is they allow their users to create their own avatars. But to create their own avatars and to do it in a way that's really appealing to users is graphically intense. So they run the Eve Online avatar creator on top of AppStream. And it drives more engagement for their users and they tend to stay more if they actually feel some more personal connection to who their being is in this world, in this new world. TrueGolf is another good example. So TrueGolf is the leading golf swing simulator in the world. And typically, if you want to use one of these, you have to buy a very expensive, typically about $10,000 simulator, swing simulator. And because they're so expensive, most normal human beings don't have them in their homes or where they want to use them. You have to go to a golf shop or a golf store. And so what TrueGolf wanted to do, they faced the same dilemma I was talking about earlier, which was they wanted to reach a lot more people with their software. And so they use AppStream to stream that swing simulator. And so now all you have to do is buy an inexpensive swing sensor. And then they're running through AppStream, the swing simulator, it streams down to their tablet, and you can either look at it on the tablet or you can then connect it to your big screen and see it there so it completely changes who they can reach. Here's another example of a brand new service that we launched at ReInvent, which is Amazon Workspaces. And this was another problem we were trying to solve on behalf of our customers. So anybody that's managed desktops on-premise knows that it's tricky and it's a little bit stressful, both because it's hard to manage and because the security concerns of having it so distributed are hard to deal with. And that's why you saw the rise of the VDI space, because the promise of having centrally managed desktops were very attractive from a manageability and security perspective. But the problem was that these VDI solutions forced you as customers to manage all the hardware, manage all the software, manage all the data center and the infrastructure pieces. It was very expensive and the performance was flaky. And so over the last couple years, I would say unaided. This was the request we got most frequently from companies. Could you please, please, please do something in the desktop space, the way you have for the other AWS services and platform you've built. And so that's why we built and launched Amazon Workspaces. And it gives you those benefits of central management and security, but you don't have to manage the hardware. You don't have to manage the software. You don't have to manage the infrastructure or the data center piece. The performance is really strong and consistent. And it's half the price of what the typical VDI solutions are out there offer. Now, we've announced this in limited preview at Reinvent in November. And over the last few months, we've had over 10,000 customers sign up for the limited preview. And many, many have used and given us great feedback that we've been able to build back into the product. And I'm excited to announce that today, we're making Workspaces available for everybody. And with that, I would like to bring up to give you a quick demo, the general manager of Amazon Workspaces, Gene Ferrell. Good morning. So as Andy talked about at Reinvent, we announced a preview of Amazon Workspaces. And today, we're making available to all of our customers. Now, for those of you that aren't super familiar, Amazon Workspaces is our virtual cloud desktop that's designed to allow users to connect their workspace to an on-premise active directory and give them access to their apps, documents, and other resources from the device of their choice, including PCs, Macs, Android tablets, including Kindle, and the iPad. Today, what we'd like to do is show you just how easy it is to provision, use, and manage Amazon Workspaces through the AWS console. So I'd like to introduce Paul Duffy, who's going to be assisting me with this presentation. Paul's our marketing manager on Workspaces. And why don't we start with provisioning, Paul? Thanks, Gene. Good morning. So we're going to start this demo off in the Workspaces Management Console. And you can see we've got a number of Workspaces that are already provisioned. Each of these Workspaces is a virtual machine that's dedicated to one of those end users that they can use to access, with the choice of their device, their applications, data, and other resources. To launch a new Workspace is pretty simple. We'll start with the Launch Workspaces button. The first question we're asked is which directory do we want to use to select users from? Now, this is important because the directory that we're selecting here is an existing Active Directory that we've integrated with using a simple process with the Workspaces service. So this means we're going to pick users who already exist in a customer's Active Directory who will have those Workspaces provisioned for them. So the end users can keep using their existing credentials. So the next thing we do is search for the users that we want to provide a Workspace for. Andy sounded quite enthusiastic, so we'll give him a Workspace and also add one for my colleague, Deepak. So once we've searched for those users and added them, the next thing we get to do is choose the bundle that we assign to them. Now, a bundle is a collection of hardware and software resources. We have two families, standard and performance. Performance bundles gives more CPU memory and storage resources. And in each of these families, we also have a plus option that comes with more software pre-installed. You can choose to install software in your Workspaces whenever you like, and this gives you the option of starting with a bit more software. As Andy's ultimately my boss, we'll give him a performance plus Workspace and Deepak can go with a standard one at $35 per Workspace per month. We review this information and then with one click, we start the provisioning process for those Workspaces, which will take about 20 minutes to complete. So when your Workspace is ready, the user will get an email with all the information they need to get started. The first step is to download the client of their choice from the link in the email, then install it on the device. Once it's installed, they'll need to register that Workspace with their directory, and they'll do that through the registration code, also included in the email. The final step is to then log into their Workspace using that same familiar username and password they use to access their on-premise resources. Paul's now demonstrating that using a Mac. So we've just connected to my desktop. As this is based on a performance plus bundle, we can see at the bottom of the screen that we already have some applications pre-installed, Microsoft Office in this case. So we can start up this Word document and see the nice graphics in there. We can choose to start with a web browser and browse the Workspace's home page or look at anywhere else we want to look at on the internet. Or we can also start Outlook. Now, one of the things that's really interesting about this, it seems I'm just opening a mailbox, but from a user's perspective, this is connecting to an existing exchange server. It's using my existing Active Directory credentials, and it's giving me the same email, the same experience that I would have with any other desktop. I don't have to change my behavior. Looks like I've got an email from Andy here who wants a workspace, so we can say, check your email, and he'll have one at the end of here. We're going to go back to the Word document, and this graphic looks like it could do with a little bit of rework, so I'm just going to add a comment in here. So while Paul's doing that, when we were building Workspaces, we heard a lot of feedback from customers about how important it was to support today's mobile workforce. These are users that move from device to device throughout their day, depending on where they are and what their context is. To do this, we made Workspaces persistent, so the session, the app state, all your docs persist between sessions and across devices, so you can go from different situations and always have your resources with you. So to demonstrate that, Paul's going to move from his Mac to his iPad. So you can see we've switched over to the iPad screen. He's just launched the Workspaces app. This is really important for a user when you think about a typical day in the life. You know, you might be at your desktop working on a presentation, following up on something, and then be called off to a meeting. Rather than worrying about packing up your laptop, making sure everything is synced between devices, now you can simply grab your tablet, head off to your meeting, and have confidence that when you get there, everything that you were working on at your desk is now with you in that meeting. So here you can see the iPad has come back. Same document it was a minute ago. So we're using an iPad Air, not this one, the one that's in front of me now, to access this workspace. And you can see because of the persistence features that Gene talked about before, we're left exactly where we were. The last thing I did in this Word document was add the comment. I'm straight back into that Word document. I could go back into my Outlook inbox. One of the other things we want to highlight here, the desktop experience wasn't originally designed for touch devices. So as we built the service, we added a number of capabilities to make it easier for users to have the best quality experience on a touch device. They can bring up a keyboard on screen if they want to type. We also have this slide-out radial control that gives them access to things like changing the mouse mode or using the Windows key if they need to get the best interaction with that Windows experience. My personal favorite, one of my favorite applications, MS Paint, so we can see whether my drawing skills have improved over the last few years. Clearly Paul's art skills need a little bit more work. Now let's take a look at it from the administrator's perspective. If we can bring up the administrator console. So one of the things we tried to focus on with workspaces is the IT administrator, that Windows administrator. And we wanted to give them the ability to use the same familiar tools that they use today to manage desktops in an on-premise environment to manage their cloud resources through workspaces. So when we're connected to an on-premise directory, a workspace behaves just like a traditional desktop would in the enterprise network. So what was the idea for those of you on Windows administrators is the Active Directory Users and Computers Management tool. And it's a standard Windows tool. There are a bunch of users here, but the more important thing to show here, in this organizational unit here I've called engineering team workspaces. I see these computer objects. These represent workspaces that have joined an existing Active Directory domain, which means from the point of view of me as an administrator, I can use existing tools to connect to and manage these workstations. That might mean something as simple as collecting to a Windows event log. It might mean using system center for application deployment or whatever software tool I'm using for that, deploying my kind of anti-virus software, deploying group policy. Users get the benefit here because they can continue to use their existing Active Directory credentials to access their resources seamlessly like you saw with Exchange before. Administrators benefit because they can manage these desktops just like any other desktop in their organization but without that heavy lifting of provisioning as you saw at the start. So no upfront investment in hardware or software, no complicated new training for your IT administrator to learn to be able to manage workspaces. And with our simple pay-as-you-go pricing, you can provision from one to thousands of workspaces and only pay for what you use. No long-term commitments, no minimums. We're really excited to be bringing workspaces to everybody today and we hope that all of you take the chance to sign up. Thank you for your time. Eugene and Paul, we're very excited about workspaces and our customers who have been in a limited preview are very excited as well. So we talked about the first three reasons why people are moving so quickly to the cloud, the improved agility, the very broad platform breadth and capabilities, and then the fact that your infrastructure platform is continuing to evolve with little effort from yourself. The fourth reason that people are moving so quickly to the cloud has to do with cost savings and cost flexibility. And so the first thing is you get to trade capital expense for variable expense. Instead of having to lay out all of that capital upfront for servers or data centers, before you know what the usage is going to be, instead you get to pay for as you consume it on a variable basis. So you get to turn CapEx to variable expense and then that variable expense is lower than what most companies can do on their own and that's because we take our very, very large scale that we've built in AWS and pass that to customers in the form of lower prices. And we've lowered our prices on 41 different occasions over the last half dozen years, largely in the absence of any competitive pressure to do so because we know our customers want to continue to have lower and lower prices to operate their business and to do more. So you get to trade CapEx for variable expense, lower variable expense than you can do on your own. The third thing is that there's a number of pricing models that give companies of all sizes and all situations flexibility. So if you just want to consume AWS by the hour, for instance, with EC2, common go as you like, people use on-demand instances. For those with steady-state workloads who want to make a small commitment upfront of some money upfront in exchange for hourly prices that are 50 to 75% lower than what you get in on-demand, people use reserved instances. For those that actually have opportunistic workloads that can afford to be interrupted and where they want the lowest possible price, they often use our spot instances. And so what we've done with the spot instances is we take all of our excess capacity at any one moment and we make them available in a spot market and customers bid a max price they're willing to pay to run that hour. And if they're over the market clearing price, then they get to run, and if they're not, they're not. So this is a way, if you have opportunistic workloads that can afford to be interrupted and get done over a period of time to get even better savings. And what we find is that most of our large customers, for their steady-state workloads, they run reserved instances because they're saving so much on the per hourly basis. And then they fill in the rest of their workloads with either on-demand or spot, depending on whether or not those applications can afford to get done over a period of time and be interrupted. And then we have lots of opportunities that the more you spend, the more you save. So we have tiered pricing, we have volume discounts, and then for customers that are doing very large movements of their applications, AWS, we do custom pricing deals. And so I thought I would share with you our 42nd price reduction. This will be effective April 1st, and it's a few different services. First, on the S3 side, we're lowering our prices on average about 51 percent, and the pricing tiers will range from 36 to 65 percent, depending on the tier. And that'll be worldwide. On the EC2 side, we're lowering the price of a number of our instance families, so 38 percent for the M3 family, 30 percent for the C3 family, and then 10 to 40 percent across our previous generation of instances, M1s and M2s and C1s and CC2s. And these are the Linux prices, but there'll be comparable types of price decreases for Windows and other operating systems, and you'll see that for on demand and for reserved instances, and again, across all the regions that we operate. Then our relational database service, RDS, we're reducing prices on average, 28 percent. Both single AZ and multi AZ across all the regions and across all the flavors of RDS, MySQL, Postgres, and Oracle. SQL Server will have comparable discounts as well. And then for our caching service, Elastic Cache on average, the prices are declining 34 percent. And then our Elastic MapReduce service, we're lowering prices 27 to 61 percent on the instance, almost all the instances are having their prices decreased, depending on the instance type. We're lowering them from 27 to 61 percent. We also are adding at this time support for additional instance types that you can run with EMR. So you'll have M3s, C3s, I2s, CR1s, and G2 instance types available to run with Elastic MapReduce. So for those who have been AWS customers for a long time, or those who follow the company for a long time, lowering prices is not new for us. It's something that we do on a regular basis. And whenever we can take cost out of our own cost structure, we will give them back to our customers in the form of lower prices. This is something we've done now 42 times over the last half dozen years. And I think you can and should expect us to continue to do this periodically as we can take cost out of our own cost structure. These are just a few examples of customers and how they think about cost structure in the cloud. So this is mortar and their comment was without AWS, we would have needed to raise $7.1 million to launch instead of $1.8 million. So it really changes your ability to get further with your idea and see if you have something meaningful. Dow Jones migrating data centers to AWS contributes to a global savings of $100 million in infrastructure costs. Commonwealth Bank of Australia, we've have storage costs. We're looking forward to saving hundreds and millions of dollars. Or the US Navy, which is moving a lot of workloads to AWS as we speak. And they said the move could save the Navy as much as 60% versus the cost of hosting or managing data in their own data centers. So very substantial cost savings in the cloud relative to doing it on-prem. And to give you an idea of how these four benefits that I mentioned and reasons that people are moving so fast to the cloud work for them as well as how they think about their cost structure in the cloud is my pleasure to introduce the head of operations at Flipboard, Joey Parsons. Good morning, everyone. My name is Joey Parsons and I'm the head of operations at Flipboard where our mission is to build the world's best personal magazine with over 100 million readers and a quarter of a million being added every day. We use Amazon web servers to deliver a fast, beautiful experience to you. For those of you who are unfamiliar with Flipboard, we bring a multitude of amazing content in a flippable format to your iOS devices, Android tablets and phones, Windows tablets and web browsers, and do this by curating the best of the web, your favorite RSS feeds and blogs, content from your social network, whether it be beautiful pictures your friends are posting, interesting links they're sharing, unique content they're creating, and other activity within your social graph. We also work with over 200 of the best digital and magazine publishers worldwide to bring the best seamless news and reading experience beautifully and directly to you. Just one year ago, we launched the ability for you to create and share amazing magazines for things that you're passionate about and engage at a whole new level with your friends, readers and followers. It really enables you, the reader, to become the editor and build immersive experiences within Flipboard. We've had over 7 million magazines created in just about every category and topic you can imagine. It's impressive to search our magazines and find exactly what you're looking for, expertly curated by a thought leader or a respected person of that world. Now, with this kind of scale and growing, we have to make important decisions about our platform every day and from the beginning, we needed a platform that could scale with us and quickly. We chose Amazon Web Services from day one after evaluating a few other cloud providers. From a cost, performance, tooling, and maturity standpoint, we naturally went with AWS. It also forced us to focus on distributed services and fault-tolerant systems from day one. One of the more important pieces of the AWS infrastructure for us is CloudFront. We make heavy use of static content, especially images, and our proprietary paginated content is served directly out of CloudFront. Whenever there's a new point of presence being brought up around the world, it helps our international reader base have a faster reading experience. Beyond CloudFront, there are two big portions of Flipboard's architecture using EC2 and other AWS services. We have our real-time interactive systems that serve reader and magazine content with real-time integrations of social activity. We also have our data importing platform that in just new publisher content RSS feeds and streams the data from your social networks. Now, the scale of both of these architectures is extreme, and with an international reader base and new content being created on the web every minute, we need tried-and-true systems to be able to meet the rigors of that demand. One key decision we made early on was to use MySQL and in turn RDS. With the auto-minor version upgrades, automatic backups, easy restores, and the ability to spin up read replicas to add capacity made it a no-brainer for us to choose what AWS provides. It allowed us to focus a little less on MySQL administration and a little more on scaling out the rest of our services. December 5th was a sad day around the world as Nelson Mandela passed, and our editorial staff quickly created an awesome magazine to commemorate his life. While we knew that there would be a ton of engagement, we had no idea what to expect next. We went from zero to millions of readers for that magazine in just a matter of minutes. Now, the story is simple and one you've heard many times before, but being able to react to a drastic change in demand and add new machines is a stark contrast to the alternative. Having your moment and disappointing your readers with poor performance or unavailability. One of the more important lessons we learn in regards to using AWS is to control our costs as our reader base grows. We have a constant need to add new machines to our environment to meet the demands of our readers' activity. A few important lessons stand out. We make heavy use of three-year reserve instances to save greatly on costs. If you look at the general benefits, you're essentially getting an instance for three years at the cost of just nine months. This graph shows, with the blue lines, our monthly active readership growing since inception, and the white line shows our cost per monthly active user during that same timeframe. So while we've been able to increase our monthly active readership month over month, we've been able to considerably drop the cost per monthly active user from 20 cents to less than one. Another way we've been able to save and perform better over time is through the frequent slashing of AWS pricing and whenever new instances get out into the fold. Whenever there's a price change, we evaluate the impact this has on our service and the assumptions that we made before. Maybe we used a medium, when it might make more sense to use a large now of a different class type and perform even better. Whenever there's a new instance type, we'll take a service roll of a similar instance type and deploy that new instance and see how it performs with a portion of our production traffic. If it holds up, is fast, and the pricing makes sense, we'll switch over an entire fleet of our service immediately and reap the benefits. The recently added C3 and M3 instances have done exactly that. We've been able to replace a significant amount of our C1s and M1s with fewer C3s and M3s simply because of the bang for your buck on processing power and the fact that they're backed by SSDs helps a ton too. We're also very excited about I2 instances and how we may be able to replace some of our age-based clusters that currently use high-end 4x larges and perform even better. Now, we're not only efficient from an operational expenditure position but also on team and staff. If you looked at our Amazon footprint and equated that to a traditional data center experience, we'd have a huge operations team at Flipboard. From network engineers managing the network, data center staff constantly racking new machines, dealing with provider and procurement relationships, we'd easily have a team of 15. We've been able to scale to over 100 million readers with just a full-time staff of three. We're incredibly happy with what we've been able to accomplish with AWS thus far and looking into the future, we're loving some of the new products that have been released recently, like Kinesis and the impact that can have on our real-time stream processing and will continue to evaluate and test DynamoDB for our interactive services as it continues to mature and grow. One recent announcement we're really excited about is we're incredibly stoked to see AWS expanding into China. China is a strategic region for us internationally as we already have a large reader base there and variants of our application servicing them. Having the ability to serve content directly out of China and build out infrastructure there helps us tap into one of the biggest markets of readership around the globe and we're excited for AWS to help us deliver. Most importantly, we continue to look forward to new instance types that help meet the needs of our demanding workloads and the constant incremental improvements that are being done to AWS that help continue the story of flipboard readers having a fast, beautiful, and always available experience. Thank you. Thanks, Joey. Very neat. So a question that we get asked quite a bit is how are enterprises thinking about the cloud now in 2014? And the way the question gets asked is people say, what are some of the patterns or the macro strategies that enterprises are using to think about and actually move to the cloud? So there are about seven that we see frequently that I thought I would quickly identify for you. The first is whenever there's any huge shift like the cloud is, the first workloads to go are development and test and you're seeing that in the cloud as well and there are lots of examples of it. You can see Lionsgate, the entertainment producer is running SharePoint and SAP workloads and DevTests and AWS on their way to production in AWS. The Tokyo Stock Exchange is running a variety of Oracle workloads and DevTests on top of AWS and Galata Chemicals, the added as producer in Connecticut is running their SAP DevTests on top of us and then saving 70% of the process. So DevTests are typically the first workloads to go along with altogether new workloads and that makes sense because there's no legacy. There's no migration. It's kind of a no-brainer. You get a better cost structure, better agility. You don't have to do all the work of building all the components yourself. You get the fault tolerance and multiple availability zones and regions at your disposal and you see that here too. You can look at PBS that has built all their video streaming apps to work on AWS or you can look at Quantist Airlines in Australia so they wanted to build a hotel booking website and so they built it on AWS. It's called huru.com and this is a supplement to their existing airline business and it's been very successful for them. Or GE has built a manufacturing collaboration application which allows GE engineers to collaborate on materials and models and simulation and equipment in ITAR compliant environment which was important for them given these products. So you see a lot of altogether new workloads moving to the cloud first along with DevTests. Then there's a couple variants of hybrid workloads that we've seen more and more of over the last 18 to 20 months. So first you see applications that were built on-premise often either at the same time AWS was really expanding or before they want to tap into resources in the cloud. So you can take the NASDAQ MarketWatch application. This was built right around the time that AWS was launching and they have a lot of analysis that they'd like to do during the day but the amount of data is so much that they can actually process that they don't want to actually add all those servers and data center capacity. They want to use the cloud to do that analysis and so they use Elastic MapReduce to do that multiple times during the day and then it comes back to the application and users can use it multiple times a day. Or you can look at SunPower which is a global solar and energy provider and they have a leasing business and a leasing app that they use AWS for disaster recovery and that has allowed them to have disaster recovery reliable disaster recovery in a much shorter timeframe than what they could do on-prem or in a colo and they save $120,000 a year versus running that in a colo. Or Nokia which has a very large mobile business and not surprisingly they want to do a lot of analytics on those mobile users and they had a data warehouse solution that was expensive and hard to manage and fragile and they moved to Amazon Redshift or data warehouse service and they're now able to run those same queries they run twice as fast for half the cost. So big difference. So you see workers that were built on-premise leveraging the cloud and you see it the other way around where you have applications that were built in AWS that are leveraging on-premise assets. So Samsung is an example of this they run their smart hub content on top of AWS but they want to use their transaction processing system that lives on-premise. So when people go to make a purchase of whatever content they're buying they move over on-prem to run on that transaction processing system and then seamlessly go back to the application that's operating in AWS. So you see movement both ways in the hybrid world. And then the next couple are really about migration and we're seeing an enormous amount of migration from enterprises to AWS over the last two years. You see it at the application layer like this and you can see here are a couple examples. Unilever as an example has a thousand websites over a thousand websites that are operating on top of AWS at this point 500 of which have been migrated and so they have built a template and a way to do that that allows them to do it rather quickly and easily and they're able to get new product websites out now using AWS in two days versus what took them two weeks before. Or Bristol Myers Squibb they moved their simulations and their lab to AWS for their clinical trials and what they found is that simulations that used to take 60 hours now take 72 minutes on AWS and they're 64% less expensive on AWS than what they were on-premise which means they not only save money they get more work done more quickly and as importantly it means they need less human samples which is goodness since a lot of those human samples are children. So you see lots of enterprises moving workloads migrating workloads to AWS and then you're starting to see quite a bit of entire data center consolidation and migration from on-premise AWS and this is an example News Corp has thousands of servers instances running on AWS with thousands more moving they have a pretty substantial initiative to reduce their data center footprint from 40 down to six and they're consolidating and migrating a lot of that to AWS and it's not only allowing a lot more collaboration but they also expect to be able to save a hundred million dollars in the process and then the last of the seven patterns that we see in enterprise movement in the cloud is the all-in strategy and this was made famous I guess by Netflix and Reed Hastings several years ago and what Netflix figured out was that they didn't want to have to continue to be world-class at operating the infrastructure which was important to their application as it is to any application but it didn't differentiate their business in any meaningful way and they were actually quite good at running the infrastructure we work closely with them and they're very impressive but they didn't want to have to keep investing in people and tools and resources to have to operate that infrastructure themselves they wanted to be world-class and the best anywhere in providing great original content and viewing experiences and so what you saw was over a couple year period Netflix moved virtually all their applications to AWS and you're seeing the same thing with a bunch of other companies the Kempinski hotel group in Germany is doing the same thing Suncorp which is the very large insurance financial services company Australia announced at our conference in November at Reed event that they're in the process of moving all of their workloads to AWS and so when Reed and Netflix announced this several years ago was pretty pioneering I would say today it's not as pioneering you're seeing more and more companies move all into the cloud and if you do so the advantages are you actually get to enjoy all the benefits of the cloud faster you figure out how to operate in the cloud faster you figure out how to build value where you can really differentiate your mission your business faster so you're seeing more and more of this you're going to see more announcements throughout the year of these types of movements to AWS I'd like to now bring up an enterprise that is very forward-leaning very innovative and is blazing new trials going all in in the cloud as well it's my honor to introduce the CEO of N4 Charles Phillips Good morning I'm Charles Phillips CEO of N4 and we always like to start off with a quick little video intro so great just to complete the introduction to N4 we're the third largest business applications company in the world you probably hear the term ERP and we build ERP applications for specific industries and it's really a transform company really over the last three and a half years we've transformed it to something completely different if you do N4 before a bit of context I came in December of 2010 with three colleagues from Oracle and we all joined the same day and we spent about a billion dollars over the last three and a half years doing one thing building great products with beautiful user interfaces so that's part of our differentiation we're the only applications company that has a separate creative agency in New York City and that's where we're headquartered and we did that because we thought user experience was going to be a big differentiator for us going forward the next generations of users would demand more intuitive more consumer-like interfaces and so we're taking that skillset and bringing it to the enterprise so we have a separate agency and we've hired people from media companies advertising agencies we have a full surprise winner for information graphics we have the digital effects artists who did the movie The Avengers we have Kenneth Cole designers these are people you can find in New York that went there for creative reasons and ended up somewhere along the way picking up software skillsets we call them left-brain creatives so that's one big part of our differentiation the other thing we do differently is build applications by industry so all of those last-mouth features that normally people don't like to do because they're hard and leave to integrators and consultants to do we've made that part of the core product and so after all that investment over the last three and a half years 11,000 new enhancements by industry we delivered 1,000 integrations really reinventing the product and becoming a product-focused company we're now at a point where it makes sense to start to pivot and go to our next disruption which is going to be cloud with Amazon so along the way we made some investments more tactically we started moving customers to the cloud we have about 2,000 customers that are there now about 12 million consumers who use those applications but that was more on an opportunistic basis but it's growing so what's changing now is there's a big opportunity in the enterprise industry software industry to really move customers more aggressively because we think they're ready and what's changing is in the old days they were more concerned about the public cloud not being secure we had these conversations long conversations with customers and our user groups about I'm not ready because I'm in the airspace and defense we've secured a big deal to us or I'm in public sector these are all big customers for us but that conversation is changing most of their security breaches are now internal and they actually view the cloud especially Amazon's cloud probably more secure than their own facilities the second thing that's changing is they realize they have to move faster and so all the industry features we've been adding allow them to actually view the software as a service because they don't have to customize anymore so they want the same continuous innovation and fast provisioning that they see these SaaS companies doing in other parts of their business then last thing we decided to do is we had to lower the cost of the infrastructure as well not just the actual cloud service and so we went open source so we spent a lot of time in the last two years moving to Postgres J Boss Linux with the first large ISV to have a completely open source stack with support that's what we're going to run on Amazon as well so as we looked at this essentially we're almost a cloud startup because we're about to move our customers 70,000 customers pretty aggressively and say we could do whichever whatever we wanted to which direction we wanted to go who do we want to partner with so we decided number one we don't want to build data centers we're a bunch of entrepreneurs and basically owner operators and every dollar has to mean something for our shareholders do we want to build data centers and buy servers and start to spend that money that way not really so we basically say a cloud 2.0 company you wouldn't build a data center friends don't let friends build data centers anymore it's no reason to do that and so that's we want to build great applications not data centers and so based on that opportunity that we see our customers are sounding differently they're saying you know I said no two years ago let's talk about it what's changing is in the enterprise they started using services like Amazon mostly for dev and testing custom applications things that they thought were kind of low risk and then they migrated on to what we call edge applications things that aren't mission critical but around the edges so HCM applications we have a couple thousand customers abroad HCM suite Sierra we have marketing applications so that got more comfortable with that but the core ERP for things like financials and manufacturing that's largely still on premise but people are asking the question now should we start to move this it hasn't been available most of the companies that provide these applications are still largely on premise and they prefer it that way we're private if there's an opportunity to actually disrupt the market and make a shift we can actually afford to do that and then what's even less likely to be in the cloud is the core industry that last malfunctionality I talked about so our patient classification system for hospital or formula management for beverage manufacturer the core mission critical thing they care about they kind of were comfortable with that on premise at least historically so what's changing is they're now saying if there's a way to move all of this at once where it's all integrated same user experience and it's someone I trust who've been running these applications for me in the past if they'll go to the cloud with me I'll consider it so that's exactly what we're announcing today we're going to do is we're going to hopefully disrupt the industry and we're introducing something called in for a cloud suite the world's first entering industry suites but in the cloud and we're going to use amazon exclusively for this so all those last mile features I talked about that we put into the product this actually was a pretty good timing because you need to do that to make it a standard service by industry we also introduced something called upgrade x where you can pay basically $55,000 we'll convert your data and move you to the cloud for existing customers who've been loyal to us and have our partners doing the same thing and of course the beautiful UIs were known for all that on AWS we have customers in 164 countries it's one of the few companies that could scale with us so we'll have micro vertical cloud suites we'll call them by industry they'll be a cloud suite corporate a cloud suite healthcare a cloud suite aerospace and defense automotive so on and so on this is the way we think of clouds because this is where our customers deploy applications by industry and so to give you an example to click down on one of them in for a cloud suite healthcare this is what it would look like you'd have all the edge applications I described for financials HGM systems onboarding all the things you'd expect in the back office but then a lot of things that are unique to healthcare like clinical integration with the number one in clinical integration hospitals by the way we run if you look at the hospitals in the U.S. with more than 150 beds 72% of them run in for applications we'll have other things like nurse scheduling surgical instrument management patient acuity all these things that are mission critical for a hospital these things must work all the time we have those and so we've integrated those in since they know we understand their processes we've worked with them a long time we can now say it's okay to move Amazon has a great reputation and brand name they're HIPAA compliant all the security requirements for that industry and it's an industry under pressure right now their cost model is changing they have to go to population management away from fee for service they need to know their cost they need best practices and so automation is a big answer for that and so the timing is great to work with Amazon on this so we think Amazon is the perfect partner for us they have a global scale we need it we're in a lot of different countries a lot of localizations we've already built for different countries we had to change a lot of our products and make an investment to do that we wanted to take advantage for things like auto scaling to do that we had to expose state in our application make them stateless that way when the session ends we can give that capacity back to Amazon and not pay for it anymore so things like that require investment it's hard to do that across multiple suppliers we decided to pick one and go deep and make sure we optimize all the patching and management tools around that and all of these services that really make it a cloud service weren't available on all the other platforms and then lastly just the cultural aspect the fact that Amazon and Andy and his team were so responsive and they acted like entrepreneurs like we are and we were able to make these decisions and develop these applications or work closely with them and making happen instead of sitting around waiting for meetings that took months with some of the other guys this was happening in days and weeks so we just knew we could work with them and so a very responsive team customer focus and it's the only service provided we've ever worked with that actually called us to help us lower prices voluntarily so we just liked the breadth and the depth and the culture that they were leading to us too in the innovation so we're essentially going all in on AWS moving our customers there we're announcing this today we'll be doing a series of webcasts and events around the world with customers the same thing with our partners we're changing the incentive compensation plans for our field organization and our partners to encourage them to move people to this platform we think it's the right one for us and it was had so much more capacity than anybody else out there so we're looking forward to being a great partner for AWS and hopefully it will stand good things about us here next year thank you thank you Charles and that's my favorite line of the entire day maybe the year so far that friends don't let friends build data centers it's very good we're really excited about partnering with Infor for many years to come so this is an oldie but goodie topic it allows our audience to have a perspective that's balanced that it's not just you know the vendors talking to them it's the community it's analysts it's technologists it's customers practitioners so they get a full perspective that's unfiltered it's interesting the the TED conference the $8,000 ticket annual conference used to be amazing material amazing talks 18 minutes long but you could only see it if you bought a ticket then they started making the talks available online and the argument was well then who's gonna buy a ticket that was like a giant ad for going to the actual event it's like it's like who would go to a