Amazon’s announcements at re:Invent, its 2017 developer conference in Las Vegas last week, define the beginning of new era in IT. The company announced a raft of new managed services that enable customers to focus on extracting business value from their application investments without the demands and costs of designing, building and operating complex infrastructures. Central to this third wave of cloud computing are new managed services for serverless-computing, machine learning, data analytics and the Internet of Things.
Amazon is executing a strategy to maximize time to value. The third wave of cloud computing will enable customers to focus on the business logic of their applications, derive analytical insights from real-time data streams and leverage the predictive power of deep-learning models, all while leaving the infrastructural complexity and operational heavy lifting to Amazon.
Businesses that have not already made the move to the cloud face increasingly stark choices: either leapfrog to third-wave managed services or be at a strategic and competitive disadvantage. To put it bluntly: CEOs and boards of directors who lack this understanding, or are unwilling to make these choices, are failing in their fiduciary responsibilities.
The $18 billion gorilla
That Amazon can attract 43,000 paying customers to its annual technical event underlines the company’s dominant and influential position in the cloud computing business.
Amazon’s cloud business dates to 2002 but was formally relaunched as Amazon Web Services (AWS) in 2006. The AWS business now generates over $18 billion in annual revenue for Amazon and—despite aggressive competition from Microsoft and Google—AWS’s market share grew from 39 percent to 44 percent year-over-year. In the last five years AWS have launched 3,951 new services and features.
The pace of innovation in the AWS business ensures that Amazon remains not just the market leader but also the thought leader in cloud computing. Amazon’s announcements of new services and features at re:Invent every year have a habit of defining the agenda for the cloud business and setting the benchmark for others to aspire to. This year’s conference was no different.
Waves of cloud computing
The cloud business, since its inception in 2006, can be divided into two distinct waves. Early cloud computing capabilities enabled moving existing application workloads from private data centers to utility infrastructure managed by Amazon. Customers gained the advantage of reducing capital investments in complex physical data center, while tapping the cloud’s ability to provide utility infrastructure on demand, with costs based on consumption.
Most applications moved during this first wave could not take advantage of the native scaling and resiliency of the cloud, but a second wave of migration emerged as customers built new “cloud native” applications composed from a growing portfolio of cloud services available from AWS and other cloud providers. These new native applications were architected to deliver the attendant advantage of cloud computing’s scaling, resiliency, management and security capabilities.
Leveraging the dynamic scaling and resilient fault tolerance advantages of the cloud environment still requires customer’s architects and developers to configure clusters of servers, wire together complex stacks of inter-connected software, implement complex networking topologies and configure and manage multiple layers of data storage technology. To deliver all of this, customers still need to hire and retain advanced architecture design, development and operations skills.
Ultimately, all of the inherent complexity of building cloud applications has limited the impact of cloud computing to companies that can afford the human capital investments or the financial investment required to pay someone else to do the work.
The third wave of cloud computing will remove the need for customers to concern themselves with the architectural and operational complexity of the computing and data infrastructures that underpin their applications. All of that complexity will be neatly packaged and abstracted away through a set of managed services offered by Amazon and other cloud providers.
As Amazon Chief Technology Officer Werner Vogels put it: “The only code you will write in the future is business logic. Everything else will be managed services.”
Perhaps most indicative of Amazon’s third-wave announcements was Amazon Sagemaker. Testing, deploying and leveraging deep-learning models is an incredibly complex undertaking. SageMaker provides data-scientists with an end-to-end suite of tools to develop and deploy sophisticated deep-learning models without having to deal with complex infrastructure issues.
Amazon is utilizing its deep-learning capabilities to optimize its own services and to deliver a growing portfolio of deep-learning enabled services for including object recognition in video, voice transcription and translation and natural language analysis.
The natural language analysis service, AWS Comprehend, also is indicative of the nature of third wave cloud services. Comprehend provides a fully managed, deep-learning enabled, natural language analysis capability delivering entity extraction, key phrase identification, source language identification, sentiment analysis and topic modeling. All of this can be done on real-time feeds of textual data and does not require the building, deployment or operational management of complex infrastructures by the customer. Businesses can focus on leveraging the analytical power of the service while Amazon ensures it scales, is reliable and secure. Customers only pay for the resources used in per-second billing increments.
Amazon’s investments in a new serverless-computing model are anchored around AWS Lambda. This new approach to application development removes the need for customers to configure or deploy server resources. Lambda, which is growing in adoption by 300 percent year over year, is a managed service that automatically executes software functions when triggered by events. Those events could be anything: from a user clicking on a button on a web page to a smart sensor sending an updated piece of data. AWS Lambda is rapidly becoming the hub for a wide range of Amazon services. A number of sophisticated cloud applications can now be composed from existing AWS services, linked and coordinated by the execution of Lambda functions.
Amazon also made a number of key announcements in the data storage and management domain that underlines the move to a third-wave managed cloud computing model. Amazon Aurora is the company’s cloud-native relational database. Aurora is one of the fastest growing services in AWS history with 250 percent year-over-year growth. At re:Invent, Amazon announced Aurora Serverless which provides database startup, shutdown and auto-scaling based on application demands, without customers having to manage any database instances or infrastructure. Customers are billed only for the resources utilized while the database is running.
Amazon DynamoDB is the company’s cloud-scale NoSQL database for unstructured data management. The company announced DynamoDB Global Tables last week, which is a new fully managed service providing automatic data synchronization and fault resilience across a globally deployed database infrastructure. This new capability removes the need for customers to configure complex replication and data distribution topologies and infrastructures. Further reducing operational complexity, Amazon announced DynamoDB Backup and Restore providing on-demand and continuous backup, with point-in-time restoration of DynamoDB tables.
One announcement of particular interest to enterprises will be Alexa for Business (AfB). Amazon is enabling its hyper-popular voice-activated smart agent technology for the office environment. AfB will integrate with common workplace technologies to provide full automation of meeting rooms, audio-visual presentations, meeting scheduling and other tasks. The AfB service will enable business customers to deploy and manage Alexa devices at scale and will enable individuals to integrate their personal and business Alexa profiles while at the office.
Other significant announcements at re:Invent 2017 included several related to the Internet of Things, new ways to manage computing instances and new types of compute instances optimized for deep learning and other computationally intensive applications, a new graph database for relationship based data topologies called Amazon Neptune and AWS Cloud 9, a new integrated environment for developers to enable rapid construction, testing and deployment of AWS-based applications.
Amazon’s announcements at re:Invent 2017 clearly mark a transition in the maturity of cloud computing and point the way to a fully managed third wave of cloud applications.
The transitions we’ve witnessed in the cloud computing business are analogous to those witnessed in the automobile industry, albeit over a vastly accelerated time scale. In the last 150 years, we evolved from craft-built motorized horse carriages to mass produced cars to now driverless vehicles. The first wave of cloud computing was built by hand and the second wave of cloud-enabled mass production of applications by the sewing together of component services.
Driverless vehicles abstract away all of the complexity of the driving process from the consumer while conveying them safely and comfortably from A to B. The third wave of cloud computing will do for business what driverless cars will do for motorists—sweeping away the underlying complexity that plagues business system IT while vastly improving time-to-value and the impact of investments in new data and deep-learning enabled capabilities.
Organizations face critical choices about the velocity and effectiveness of their transition to the cloud. Amazon’s announcements last week make those decisions imperative.
Companies that have already migrated their IT to the cloud must keep pace with these changes and the new third wave model or again face being at a significant strategic disadvantage. Being behind by two or even one generation of cloud technology places companies at significant competitive disadvantage: the equivalent of selling horse drawn buggies in a world of driverless cars.