eWEEK content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
LAS VEGAS—Amazon Web Services this week showed once again just how much farther ahead of the pack it is in cloud computing and how high it is setting the bar for the future—not just for cloud computing, but all of IT.
At the company’s re:Invent conference here, AWS CEO Andy Jassy and CTO Werner Vogels made 28 announcements over two days, each one seeming to top the last and eliciting cheers on several occasions from the keynote audiences, which this year totaled 32,000, a significant increase from last year’s 19,000.
Jassy and Co. had a little fun while they were at it. After having endured a barrage of mockery from Oracle Chairman Larry Ellison at the Open World conference in September, AWS returned the favor.
Jassy said AWS customers have the power to see through the “hand waving and bombast” of “old guard leaders” such as Ellison, who had taken AWS to task for being behind the times in its database and platform technology. Oracle’s assertions, Jassy said, were based on “wild claims and benchmarks that suit their purposes.”
AWS now can stake its own claim for equal if not superior data management technology. Jassy announced a PostgreSQL-compatible version of its Aurora database to go along with the MySQL-compatible version. Also new is Athena, a SQL query tool for Amazon’s ubiquitous storage service, S3.
Vogels added to that announcements for Amazon Pinpoint, a “data engagement service” that sounds a lot like omnichannel marketing analytics tools from Oracle and Salesforce.com. To round out the data management news, Vogels announced AWS Glue, a fully managed data catalog and extract, transform and load (ETL) service that enables users to move data between databases for more efficient analytics.
AWS also beefed up the next versions of its data transfer appliance to help users get data into the cloud. Snowball, announced last year, was so popular that AWS had to up the ante, Jassy said. The new version, Snowball Edge, can carry up to 100TB and is rack-mountable.
The next step up from there is Snowmobile, which can hold up to 100 petabytes (PB) and comes complete with its own ruggedized, climate-controlled shipping container that was hauled into the keynote area at the end of Jassy’s talk.
Going Serverless
But the big stories at the conference—and going forward—were serverless computing and artificial intelligence (AI) services.
Serverless computing, which AWS introduced with its Lambda event-driven service two years ago, appears to be poised to take over as the primary method of running compute services in AWS, in addition to EC2 instances and containers.
Several large customers, including iRobot and FINRA (Financial Industry Regulatory Authority), told their stories of converting much of their infrastructure to Lambda functions. FINRA processes 500 billion transaction validations a day, running them as Lambda functions.
So, the disappearing data center also could lead to the disappearing server. During one session AWS speakers actually brought out a magician to make a real server disappear like the rabbit in the hat. Does such a model—wherein compute is so much cheaper for the user—end up hurting AWS’ ability to make money on compute?
Not so, said Michael Liebow of AWS partner Accenture, which is in the process of converting its applications to serverless. “Serverless is orders of magnitude cheaper and accelerates the server,” said Liebow, global managing director for Accenture Cloud Platform, which runs on the AWS cloud. “This is the post-infrastructure era. It changes how you design solutions. You pay for what you use. For AWS, it’s really a volume play. The usage scales up even faster [than running on EC2].”
Accenture’s AWS Business Group, for its part, announced an expanded relationship with AWS around internet of things systems, transformation services, and analytics and big data Services.
At re:Invent, AWS announced three major initiatives around serverless computing. The first, AWS Greengrass, is a service that enables Lambda to be embedded on devices, giving them the ability to execute code on their own without having to send data back to the cloud (Greengrass is also included on the new Snowball Edge).
The second, Lambda@Edge, is a model by which developers can put JavaScript code at AWS edge locations to execute remote functions. Finally, AWS Step Functions is a way for developers to map out in a visual workflow format how serverless components interact with each other and in what order.
All in with AI
AWS has been doing AI for a long time, and parent company Amazon is making a lot of money with its Alexa AI-powered Echo devices for consumers. At re:Invent, Jassy announced a new set of AI services that enterprise developers can now use for their own applications.
Amazon AI includes Rekognition, a facial recognition service; Polly, a text-to-speech service; and Lex, which is the natural language technology from the Alexa device. Several customers are adding what are called Alexa “Skills” to their own applications, including Expedia and GE Appliances, while Intel is working with AWS to create reference designs for OEMs.
It’s easy to see how just about any application that can tie back to the AWS cloud will be capable of having voice-enabled and intelligent services. The next step is proactive applications, in which Lambda functions can fire commands to a mobile device or Echo speaker to remind users to take out the garbage or take their medicine.
As he did at the AWS Global Partner Summit earlier in the week, Jassy enjoined attendees to embrace the direction AWS is headed. He reflected on the 10 years of AWS existence at this, the fifth re:Invent. “We wasted a lot of time and energy wondering if people would embrace the cloud,” he said. Now, he speaks with confidence that over time all enterprise data center workloads will move to the cloud. “The next 10 years will have markedly more innovation than the first.”
As he likes to say, “Giddy up.”
Scot Petersen is a technology analyst at Ziff Brothers Investments, a private investment firm. He has an extensive background in the technology field. Prior to joining Ziff Brothers, Scot was the editorial director, Business Applications & Architecture, at TechTarget. Before that, he was the director, Editorial Operations, at Ziff Davis Enterprise. While at Ziff Davis Media, he was a writer and editor at eWEEK. No investment advice is offered in his blog. All duties are disclaimed. Scot works for a private investment firm, which may at any time invest in companies whose products are discussed in this blog, and no disclosure of securities transactions will be made.