AWS for virtually every application- AWS Online Tech Talks

Key Takeaways

foreign [Music] hello and welcome to today's Tech talk on how AWS can support virtually every application this is a foundational session where we'll cover key reasons why customers choose AWS to build run and scale a wide range of applications in the coming months we'll have Tech talks where we'll dive deeper into technical capabilities that enable customers to run their applications securely on AWS my name is pragya Pandey and I'm a principal product marketing manager on the AWS product marketing team now let's look at the agenda for today we'll start with some context on types of applications that our customers are running today and how the way these applications are built and run has evolved over the last few years next we'll dive deeper into why organizations choose AWS and how our infrastructure supports virtually any type of application we'll also share some customer examples where companies are realizing greater security efficiencies performance and cost savings by running their applications on AWS and lastly we'll go over some next steps and resources to help you move or build your applications on AWS today applications are the Cornerstone of any business many Enterprises rely heavily on applications for Erp HR management payroll systems as well as custom applications they need for their core businesses several startups and smbs are developing interactive web applications streaming platforms real-time gaming applications and e-commerce sites to address customer needs from an insurance application used to analyze claims to an iot application used for predictive maintenance businesses in every industry are powered by applications the way these applications are built and run has also evolved significantly in the last few years few years ago applications were designed in a monolithic way and deployed mostly in on-premises data centers developers had to deal with I.T and infrastructure teams that were in charge of data center environments and it could take weeks or even months to access infrastructure slowing down innovation with the cloud you have access to infrastructure On Demand with virtually infinite capacity applications are designed as microservices making application development process more agile and you can create your own virtual infrastructure and deploy your applications with just a few clicks with Technologies like containers and serverless you can quickly build modern applications that scale automatically and reduce your total cost of ownership this agility scalability and efficiency has motivated many companies to run their applications in the cloud and use hybrid cloud architectures but despite what feels like massive adoption we are still getting started analysts estimate that by 2025 51 percent of it spending will have shifted from traditional solutions to the public Cloud compared to 41 percent in 2022 that means there are so many workloads that are going to move in the coming years with a lot of innovation yet to come this is a great opportunity for you to assess your current needs and better understand the benefits of running your applications in the cloud our CEO Andy jassy mentioned in the first ever reinvent keynote in 2012 that our focus is on enabling any application to run in the cloud as you look to further Innovation with the cloud our Focus for enabling the broad range of applications you have in your organization remains true that also includes the application you will need in the future to grow and extend your business at AWS our vision is to be the best place to build and run virtually any application securely both today and tomorrow so what does it mean to support virtually every application it means we need to support the broad range of applications that you need for your business that includes Enterprise applications such as those using sap Oracle IBM and other Erp Technologies for example Siemens migrated their legacy sap system to AWS to increase agility and reduce costs it also includes high performance Computing applications running large complex simulations and deep learning workloads such as weather forecasting and genomics research as an example researchers from Dana-Farber Cancer Institute are using AWS for faster and more efficient drug Discovery experiments it also includes modern applications such as modern serverless apps container-based apps and machine learning applications one of our customers HBO Max uses event-driven architecture to scale the direct to Consumer platform customers are also thinking about low latency applications such as real-time multiplayer games or high frequency trading applications a great example is Netflix that has empowered artists to work remotely with low latency workstations for Content development also iot applications which connect and manage up to billions of devices for industrial consumer commercial and Automotive workloads as an example John Deere uses AWS iot for precision Agriculture and for improving Factory operations and with hybrid Cloud you want to take advantage of the benefits of the cloud wherever you need it including on premises and at the edge as an example Morningstar uses AWS outposts to build and run AWS services in its on-premises data centers now when we ask Business Leaders why they choose AWS to run their applications we get a lot of different answers because organizations have lots of different types of workloads today we'll focus on key reasons that customers tell us why AWS is the best place to build and run your application securely wherever you need them first AWS is designed to be the most secure and reliable Cloud so that you build and run your applications with confidence second AWS constantly innovates so you can get the best performance and value from your Cloud infrastructure turn AWS offers the most capabilities for your applications to support your unique needs and finally AWS brings the cloud capabilities to applications running anywhere with a truly consistent experience and infrastructure services and tools wherever you need them now let's dive deeper into the first area building a secure and reliable application requires careful consideration across many areas you need to ensure the infrastructure the application is running on is secure application data is always protected and transfer of information is safe organizations handling sensitive data such as personally identifier information or pii digital assets like cryptocurrency and other financial information have stringent privacy and confidentiality requirements that they need to adhere to AWS delivers on organization's needs for confidential Computing with the AWS Nitro system the foundation of a modern compute infrastructure the AWS Nitro system offers built-in security to protect hardware and software against potential security threats AWS Nitro enclaves a feature of Amazon ec2 enables organizations to create isolated compute environments to further protect and securely process highly sensitive data we talked about the built-in security with the Nitro system that helps protect data we also ensure that application data stored in our storage Services is always protected starting with identity and access management with Amazon S3 we offer the ability to block public access to all the objects at the bucket or the account level and starting in April all newly created buckets in the region will have S3 block Public Access enabled by default with Amazon elastic file system customers can use identity and access management policies to manage access to file systems data giving them the flexibility to manage multi-tenant environments in a cloud-native way securely across the storage portfolio we offer encryption for data at rest and in transit in January this year we announced that S3 will encrypt all new objects by default this change puts another security best practice into effect automatically with no impact on performance and no action required by our customers we also offer extensive auditing and monitoring to help move towards real-time auditing and response as an example we offer Amazon S3 lens to provide organizations-wide visibility into object storage in November last year we added 34 new metrics including new metrics on access control lens metrics are integrated with Cloud watch so you can view lens metrics in cloudwatch dashboards now building on to the security Foundation we deliver data resiliency capabilities for business critical applications so you can meet aggressive recovery point and time objectives whether you're just getting started looking to migrate Enterprise applications or already building on AWS we have resources to deliver data resiliency to address data events you can Implement a quick roll back with file or object versioning or with replications you can also address data recovery needs with AWS backup to Simply backup and Recovery with the fully managed policy-based service you can also backup data on your Amazon EBS volumes to Amazon S3 by taking point in time snapshots we also offer managed Disaster Recovery Service with AWS elastic disaster recovery it helps organizations recover their on-premises applications to AWS as well as recover cloud-based applications across AWS regions check your data at the network level we adopt a zero trust model that ensures that access to data should not be solely made based on network locations it requires users and systems to strongly prove their identities and trustworthiness and enforces fine grade identity based authorization rules before allowing them to access applications data and other systems and this is made simpler with the AWS verified access that we announced in preview at reinvent 2022. verified access allows administrators to build a set of policies that Define a user's ability to access each application it is built around zero trust principles and evaluates each access request in real time and only connects users to Applications they are allowed to access in addition to providing built-in security at multiple layers AWS is also committed to offer the most reliable infrastructure so you can effortlessly scale your applications without any service interruptions AWS provides the most extensive fully redundant high performance Network backbone to deliver highly responsive reliable experiences for nearly every type of workload to give you some perspective here's a customer example where the built-in security capabilities with AWS infrastructure helps protect highly sensitive data networks is a digital currency networks for banks running The Ledger card firm on AWS The Ledger platform uses AWS Nitro enclaves to perform signature verification and cryptographic re-signing of batches of transactions using AWS M10 is able to deliver a secure performant and cost-effective solutions for the digital currency Market and reliability as foundational requirements other critical considerations for Business Leaders a performance and cost decision makers and application owners need to deliver on the performance needs for their businesses to meet slas and drive business growth you also need to ensure that the infrastructure choices will continue to fuel business Innovation while managing costs top of mind more than ever in the current economic environment we enable you to get the best performance and value from your infrastructure in many ways first we strive to deliver the highest performance so that applications can scale easily run faster and be more efficient second we constantly innovate to offer the best price performance at lower costs third we offer tools and resources that can help further optimize your costs let's look at the performance gains that you can achieve with the Nitro system we already talked about the security benefits of nitro but the primary driver for building the Nitro system was performance back in 2012 we started thinking about ways to make the AWS Cloud more efficient given that so much CPU was being used for virtualization and I O the idea was to host a hypervisor of float i o and management overhead and offer bare metal instances without sacrificing security in 2013 we started working with a company called Annapurna labs when we were developing the early version of the Nitro system the Nitro system is a combination of dedicated hardware and lightweight hypervisor enabling faster Innovation and enhanced security the Nitro system delivers practically all of the compute and memory resources of the host Hardware to your instances resulting in better overall performance of your applications at re invent 2022 we announced the fifth generation of nitro and over 20 million Nitro chips have been shipped so far now the impact of the Nitro system to you can be measured with compute performance if you take current sixth generation x86 instances with the AWS Nitro system Amazon ec2 instances can deliver over 15 percent higher throughput performance on some workloads as compared to other major Cloud providers running the same CPU this is a significant performance advantage that can make your applications more efficient and our goal is not just to deliver the best performance but offer the highest performance at lower costs AWS has years of experience in building silicon just as with Nitro AWS saw the opportunity to deliver high price performance by applying a learning to building processors AWS then custom designed the AWS graviton processors to enable the best price performance for a wide variety of applications today we have over 100 graviton-based ec2 instances across our portfolio providing you with the right mix of compute memory and storage to deliver Better Price performance for your applications and now we are on to our third generation of graviton processors graviton 3 processes provide up to 25 percent better compute performance compared to graviton 2 processors they also use up to 60 percent less energy for the same performance as comparable ec2 instances which helps your organization reduce your carbon footprint we launched our first graviton 3 based instance last year c7g and two more graviton 3 based instances m7g and r7g earlier this year one example where our customer is experiencing huge performance gains with graviton is Formula One F1 had great success with graviton 2. they had already seen that graviton 2 based C6 GN instances provided them the best price performance for some of their computational fluid dynamics or cfd workloads when graviton 3 came out they tried them and found that graviton 3 was even better graviton 3 based c7g instances were up to 40 percent faster than graviton 2 for the same simulations with the improved price performance F1 found that graviton 3 based instances became the optimal choice to run all of their CFT workflows and here's another example where we are enabling customers to save costs and drive efficiencies when running their workloads on graviton-based instances DirecTV offering streaming services are able to lower costs by 20 percent improve latency by 50 and drive CPU improvements by up to 35 percent by using graviton 3. now as you continue to drive the best price performance with graviton we also wanted to bring price performance benefits to companies running machine learning workloads on AWS these ml models are becoming increasingly large and complex growing exponentially in size to give you some perspective just three years ago state of the art deep learning models included millions of parameters and today models have grown to hundreds of billions of parameters these larger and more complex models result in increasing the cost and time for training and running large models AWS custom designed the AWS inferentia and AWS trainium chips to provide high performance and cost-effective deep learning in the cloud AWS trainium helps train ml models and AWS inferentia helps with models inference that is it helps generate predictions based on inputs from ml models last year we introduced Amazon ec2 trn1 instances powered by AWS trainium these instances are well suited for large-scale distributed training of complex deep learning models across a broad set of applications such as natural language processing image recognition and generative AI applications such as image generator Tech summarizations Etc these instances deliver up to 50 percent cost to train savings over comparable Amazon ec2 instances tr11 instances offer 800 gigabytes per second of EFA networking bandwidth and we recently announced tr1 and instances that double this to 1600 gigabytes per second to deliver faster trading times for Network intensive models such as large language models or mixture of experts for efficient distributed training of large-scale models these instances are deployed in ec2 Ultra clusters that enable scaling up to 30 000 Trillium accelerators we recently announced a general availability of Amazon ec2 in two instances powered by second generation of AWS inferentia 2 accelerator if two instances are optimized to deploy 100 billion plus parameters generative AI models such as large language models and vision Transformers at scale compared to in fun instances in two instances deliver Three Times Higher compute performance four times higher accelerator memory up to four times higher throughput and up to 10 times lower latency if two instances are for you high performance at the lowest cost in the cloud for deep learning inference in two instances delivered up to 70 Better Price performance and up to fifty percent better performance per watt than comparable Amazon ec2 instances in two instances are the first inference optimized instances in Amazon ec2 to support scale out distributed inference with 192 gigabytes per second of neuron link connectivity between accelerators now another area where AWS has been a Pioneer to help organizations reduce costs is serverless computing many customers are running their workloads on a serverless services so that they can build and run applications without thinking about servers these Services also eliminate infrastructure management tasks like capacity provisioning and patching so that you can focus on writing code serverless applications start with AWS Lambda an event-driven compute service natively integrated with over 200 AWS services and software as a service applications AWS serverless Services automate high performance scaling and reduce operational overhead so you can redirect resources towards high impact projects reducing your total cost of ownership by up to 57 percent according to Deloitte you can gain further price performance with our silicon innovation managed services such as AWS Lambda running on graviton offer 40 Better Price performance now as you've seen with our services we take several measures to deliver the highest performance at lowest costs so that you get the best value from your infrastructure with the current environment organizations are often interested in how AWS approaches cost optimization at Amazon one of the leadership principles is frugality it motivates us to look at Cost implications for every design and solution we truly believe that building such a culture will result in a long lasting and sustainable model for cost optimization and efficiency we also have internal tools and resources that help us reduce costs and finally we set metrics and goals for cost optimization and track them closely we build accountability with single threaded owners who are constantly looking for ways to optimize costs and make our services and tools more efficient by bringing these cultural changes being more intentional about cost savings and having dedicated teams or individuals focused on cost optimization can help you reduce your overall costs now Beyond building a culture of cost optimization you have options and tools to further reduce your costs first for compute there are purchasing options that help you optimize cost based on your needs you can purchase AWS infrastructure on demand via savings plans or as spot instances savings plans is a flexible pricing model offering lower prices compared to on-demand pricing in exchange for a specific usage commitment for a one or three year period Amazon ec2 spot instances let you take advantage of unused ec2 capacity spot instances are available at up to 90 discount compared to on-demand prices and customers have saved over 15 billion dollars with savings plans since its launch you also have a broad set of options for compute and Storage you can take advantage of ec2's comprehensive portfolio of instances to optimize workload performance and eliminate underprovisioned resources we also offer choice of storage classes based on your performance needs and cost considerations for example The efs1 Zone storage classes can reduce storage costs by 47 percent compared to Regional EFS standard storage classes the second to improve compute efficiency you can use services like Amazon ec2 Auto scaling and ec2 Fleet to easily add or remove capacity to meet changing demand Amazon ECS and Amazon eks help build modern applications that scale automatically enabling you to pay only for resources used by your applications resulting in lower TCO you can save up to 65 percent with storage compression and deduplication to improve storage efficiency and for your windows workloads we offer the most options in the cloud for using new and existing Microsoft software licenses on AWS migrating Legacy applications running Enterprise workloads on premises can lead to significant cost savings one of our customers World Fuel Services moved their legacy windows workloads out of 20 data centers recently to cut monthly costs by 50 percent and avoided an additional 70 increase of expenditure in 2020. and lastly you can use tools such as the compute Optimizer that recommends optimal AWS resources for applications by analyzing historical utilization metrics to reduce costs and improve performance compute after Optimizer has provided over 10 billion recommendations since its launch our storage Services also offer intelligent tiering capabilities for data with unknown access patterns customers have already saved more than 750 million dollars with just Amazon S3 intelligent hearing since its launch here are some examples where customers use the purchase options and our services to reduce costs the sales for Salesforce team was able to build a scalable elastic compute infrastructure including spot instances with its remodel infrastructure it takes the company less time to process twice as much as data by lowering compute cost by more than 60 percent saving the company more than one million dollars a month Formula One used a combination of Amazon ec2 instances to reduce its cfd simulations time by 80 percent and lower the cost of running workloads by 30 percent as a result F1 can better support its strategic priorities of increasing competitiveness and producing a world-class spectacle for fans and finally Airbnb saved a significant amount by choosing the right storage classes for the data they move the data to Amazon S3 Glacier and replaced homegrown data backup system with AWS backup resulting in savings of 63.5 billion dollars in 2020. now we have talked about how the Nitro system helps to deliver performance and enhance security and fuels innovation that Innovation is just a part of a broadest and deepest capabilities we offer here at AWS we offer the most capabilities so you have access to whatever you need to build all the applications you can imagine at AWS we are customer obsessed in fact it's another one of our leadership principles and baked into our DNA we spend a lot of time talking to customers and with that comes a deep understanding that no two customers are the same we know with conviction that customers care deeply about selecting the cloud provider with the broadest and deepest set of capabilities for their applications and we are constantly thinking of ways to invent on behalf of our customers since our start in 2006 we have been continually expanding our services to support virtually any application with more than 200 fully featured services for analytics machine learning Internet of Things applications development and more for compute we have the most compute instances to meet the needs of your workload for networking AWS has the most secure and reliable Network as well as the greatest capacity Network for storage AWS offers in Industries broadest choices of storage services and deepest functionalities for our modern applications AWS pioneered serverless Computing and has the broadest range of serverless services allowing customers to offload more of their operational overhead and AWS is the most trusted way to run kubernetes with more kubernetes containing rice workloads running on AWS than any other cloud and looking further at compute we now have over 600 instances which is more than any other cloud provider each one providing a choice of processors storage networking and operating system so you can always find the right fit for your specific workloads with over 16 years of collaboration and Innovation with Intel we delivered over 400 Intel based instances supporting a broad range of workloads including a new preview on the latest Intel Xeon scalable processors AWS were the first Cloud to launch AMD processors and map based instances in fact we are the only major cloud provider to provide on-demand Mac OS based instances this Has Lifted limitations and enabled people to develop build tests and sign Apple apps on Amazon ec2 now as we think of all the capabilities organizations need for their applications Beyond ec2 we have many options for compute you have the choice to pick the layer of abstraction you're most comfortable with and we will provide the tools services and apis necessary to help build your application you can start with ec2 and have access to all the knobs you want to build a scalable architecture or you can use a container-based orchestration services or go completely serverless with Lambda and fargate where you can focus just on your application without worrying about the underlying infrastructure the serverless services allow you to build and imagine highly scalable applications hundreds of thousands of our customers are choosing serverless Solutions and benefiting from faster deployments higher productivity fewer security incidence and lower costs one of the most important decisions a business owner will make is where to store their data for their applications AWS provides the most complete set of cloud storage services for object block and file for customers to run virtually every application at the performance and cost they need when Amazon S3 launched one of the top use cases was website hosting today we see organizations of all sizes and across every industry running every imaginable workload from data Lakes to machine learning to analytics content delivery and medical research with S3 you can retrieve any amount of data from anywhere with an industry-leading 11 9 of durability for those looking for Block storage Amazon EBS offers easy to use high performance block storage at any scale you can select block storage volume types based on your data size and iops requirements optimizing your application workloads for price and performance and if you're looking for share file storage Amazon EFS allows you to build modern Cloud native applications by providing a serverless and fully elastic file system and Amazon FSX allows you to migrate your file systems to AWS with light for like capabilities whether your applications run on NetApp on tap Windows file server luster or opencfs in 2022 we announced Amazon file cache a new storage service to accelerate hybrid workloads file cache serves as a temporary high performance storage location for data stored on premises providing sub millisecond latency and high throughput and making it easier to process file data regardless of where the data is stored our storage solution will enable your you to organize data and configure fine-tune access controls to meet specific business organizational and compliance requirements all while optimizing costs we also realized that having the choices and flexibility matters for building your applications and we want to make it faster and easier for developers to build applications AWS amplifier lets front-end web and mobile developers easily build ship and host full stack applications on AWS with the flexibility to Leverage The breadth of AWS Services as use cases evolve Amazon code Catalyst a new service announced in general availability is a fully managed service that allows you to build and deliver applications faster without dealing with the devops complexities Amazon code Whisperer is a machine learning powered service that helps improve developer productivity by generating core recommendations based on your comments in nature language and code in the integrated development environment these Solutions significantly improve developer productivity and accelerate the process of application development now Beyond needing the capabilities for your apps organizations are rethinking where traditional applications live and how they can make the cloud their primary model for all applications as we discussed earlier organizations are increasingly planning to integrate hybrid and Edge delivery options in their infrastructure strategies so keeping that in mind we are looking at how we can bring the same Cloud capabilities and management experience to the edge so that you can use AWS wherever you need it why companies are increasingly deploying their applications in the cloud they may still need to run some of their mission critical applications on premises to ensure faster response times some companies also have hybrid and Edge use cases where they need to support low latency applications process data locally or meet data residency requirements and some may be in the process of migration and modernization where they need to maintain a hybrid state in such scenarios inconsistencies across environments can impact productivity or slow innovation bringing the same Cloud infrastructure and services wherever it's needed is increasingly becoming a priority for many organizations for the last few years we have been hard at work extending the AWS cloud with new services and infrastructure options to run applications closer to wherea organizations needed to where the users are their equipment is or where they capture process and make real-time decisions it starts with our regions but we have gone further to Major Metropolitan and Industrial centers and locations within 5G Networks moving to on-premises we offer Solutions running in data centers warehouses and retail locations and we have even extended into remote and many times rugged locations sometimes with limited or no connectivity to the cloud and finally to end devices running at that edge in cities factories or motive or wherever iot devices are used starting with regions AWS has the most extensive Global infrastructure with 31 regions and 99 availability zones that will continue to expand in the future to better serve our customers a region consists of typically three or more availability zones Each of which is one or more fully isolated discrete data centers we have announced plans for five more AWS regions in Canada Israel Malaysia New Zealand and Thailand and with local zones we bring compute storage database and other select AWS Services closer to large population and Industry centers so far AWS has launched 32 local zones globally and announced 21 new international local zones in January this year we launched local zones in Lagos Lima and queretaro to help customers in these areas deliver applications that require single digit millisecond latency or local data processing Mind Body is a great example of an Enterprise migration use case where locality was an important Criterion Mind Body is a software as a service company that provides cloud-based online scheduling and other business management software for the wellness and services industry like many other Enterprises they had workloads running in their existing on-premises data centers in various metropolitan areas these Services had strict interdependencies and it was hard to move it to the cloud AWS local zones provided the low latency between applications running in local zones and their on-premises data centers using Direct Connect at the local Zone this allowed Mind Body to migrate applications incrementally and greatly simplify their ongoing migration Journey to the cloud and while we have more infrastructure in more places you may still have apps that need to reside on premises and cannot move to the cloud for those apps we've created AWS outposts AWS Outpost is a family of fully managed Solutions delivering AWS infrastructure and services to virtually any on-premises or Edge location for a truly consistent hybrid experience it is available in over 65 countries and supports 14 services without post you don't have to worry about upgrading the hardware or replacing broken paths detecting that things have failed updating control planes this is all managed by AWS as The Outpost connects back to a region it also comes with the AWS API and Management console which ensures that the same tools you use in regions are available in your Outpost and to support your iot based applications AWS offers choices based on your technical capabilities customization requirements and speed to Market needs at one end of the spectrum organizations and partners can choose to build their own custom iot Solutions with our foundational services such as AWS iot Greengrass at the other end of the spectrum you can use ready to deploy Solutions built by both AWS and partners enabling you to quickly realize business outcomes and accelerate your time to Market and in the middle of the spectrum AWS provides purpose-built iot services designed to speed the time to market for specific industry use cases such as robotic orchestration and Industrial monitoring an example is iot Twin maker for creating digital replicas of real world environments let's wrap it up by going back to a great example of application where compute networking and storage come together to help you build the highest performance and scalable applications that's high performance Computing or HPC CEUs is used to tackle some of world's hardest problems including drug Discovery and genomics weather predictions models insights on social media platform and computational fluid dynamics HPC workloads Demand large-scale computations with massive amounts of data storage and high-speed networking distributed across a large number of machines HPC users also need tools that are easy to use so that they can spend more time experimenting and less on managing the infrastructure at AWS we have a complete set of HPC capabilities we bring together new instances optimized just for HPC with Intel AMD and graviton combined with elastic fabric adapter also built on the Nitro system with high speed networking we support job orchestration and cluster management and provide an enhanced compute Fabric and file systems that scale to millions of iops today we've had the opportunity to talk about why organizations choose AWS for their applications you've talked about how AWS is designed to be the most secure and reliable Cloud delivering Innovations in performance so you get the most value from your infrastructure you've seen how AWS delivers the most capabilities across infrastructure services and tools so it is faster and easier for you to build and run any application and we discussed how AWS is delivering a truly consistent experience with infrastructure services and tools wherever you need them what all this means is that AWS can securely run and build any type of application your mission critical Enterprise applications as well as your new modern and immersive applications and the ones that you're still imagining as for next steps I highly recommend visiting the AWS for every application webpage to explore resources to help with your applications and if you're looking to migrate your application to AWS check out our Cloud migration page to learn more about our migration Solutions and stay tuned for our upcoming Tech talks that dive deeper into some of the areas we discussed today thank you so much for watching this session and we'll see you in our next tech talk [Music] [Music] oh [Music]