Future of AWS hinges on AI, Lambda and on premises

Chris Moyer

AWS has so many features that it will be challenged to manage them all and figure out which ones will grow into the next big thing. It seems like Amazon is using a spray and pray approach in quite a few areas, such as artificial intelligence, machine learning and serverless computing.

The releases of Amazon Personalize, Textract and Timestream are on the top of my list of things to watch out for. I'd expect some more serverless style features, as well as a few things that are completely out of left field, such as DeepRacer.

Lambda Layers and custom runtimes open up a whole mountain of opportunities. I anticipate a lot more about these features to come out -- similar to the partnerships you saw with EC2 Amazon Machine Images -- with third-party companies that manage a runtime and layers.

One thing I really didn't hear about was how this might impact IoT and, specifically, things like Greengrass, which seems to have been a slight flop because it wasn't true Lambda for personal devices. Firecracker's open source nature makes me wonder if AWS will look to expand more on that and perhaps offer a true run Lambda on your own device type of technology that's similar to Greengrass, although I would prefer something entirely new. I'd also love to set a region of local-1 to run something on one of my Raspberry Pi devices.

On a higher level, I expect to see some major acquisitions in the near future for AWS. The cloud provider continues to expand so rapidly that it only makes sense for AWS to absorb smaller companies to keep up the pace of innovation and prevent any service terminations. Amazon doesn't kill off services like Google Cloud; instead, it adds more and more things it needs to support.

David Linthicum

I expect 2019 to be the year that AWS goes deeper into the enterprise. With the advent of Outposts, AWS has proven that the data center isn't off limits, and the cloud provider is open to dealing with enterprise pain points that were traditionally left on premises. This means the addition of several hardware appliances that AWS will sell and support, including a database integration gateway that will provide the facilitation of database sync from on premises to cloud-based databases, security appliances that provide an even greater degree of protection, and guardians on both sides of the firewall.

These moves would have two major benefits for the future of AWS. First, we need this technology. When it arrives, it should improve integration with data centers, which won't go away any time soon. Second, it will enable AWS to penetrate markets that have been dominated by large enterprise software and hardware vendors, and to compete to develop better tools and services.

Adam Bertram

It's always hard to predict cloud vendor activity, but if I were a betting man, you can't go wrong with AI. Machine learning is trendy right now, and it's just getting started.

AWS has the ability to make machine learning accessible to the masses, and as the market finds more uses for AI, AWS will continually work to provide more products around it.

George Lawton

AWS will redefine edge computing with a combination of Greengrass and an expansion of its enterprise services, such as Amazon Connect. While other competitors limit their expectations of edge to industrial automation, AWS will demonstrate that edge computing can also extend cloud management principles to on-premises tools, such as telephones and office equipment.

Over the next year, AWS partners will start to integrate robotic process automation with native services to create greater business efficiencies.

Kurt Marko

With more market share than the next four largest cloud providers combined, AWS is still acting like a hungry startup, and it introduced dozens of products and enhancements at re:Invent 2018. Two specific areas stood out from those announcements: enterprise hybrid cloud and custom-built hardware tailored to its needs. Look for AWS to redouble efforts in both areas in 2019.

A few particular items are worth watching as we embark into a new year:

• Look for AWS to expand the use of Graviton -- ARM-based -- and Inferentia -- machine learning model execution -- processors beyond its initial use in EC2 instances and SageMaker, respectively. I expect to see Graviton variants with more cores and deployed in native services to reduce costs and improve performance. Candidates for potential Gravitron inclusion include Lambda; DynamoDB; CloudFront; developer services such as CodeDeploy, CodeCommit and CodeBuild; and business applications such as WorkDocs, WorkMail and Chime. Given its secrecy, we may never hear about such deployments in services where the processor is insulated by a service layer and not directly exposed to the customer.
• AWS said Outposts, its service for on-premises implementations of native services, uses "fully managed and configurable compute and storage racks built with AWS-designed hardware," which includes its Nitro security and network hardware. While this sounds like an AWS-labeled hardware product, don't be surprised if AWS reaches an agreement with Dell to provide Dell-branded hardware for Outposts options that use VMware as the cloud software. Remember, Outposts comes in two forms: One for VMware and one with native Amazon services. If AWS and Dell work together, it could put a strain on Dell's longtime relationship with Microsoft, which has its own hybrid cloud offering in Azure Stack and a Dell-built Azure Stack rack.
• A perennial criticism of AWS and other cloud vendors is that their pricing model, particularly for infrastructure services, is far too complicated and often requires additional software help to navigate and to optimize costs. Now that AWS has an on-premises enterprise option in Outposts, look for it to introduce a simplified purchasing model. One option would be a bulk purchase of service credits that are automatically applied to whatever the customer chooses to use and where unused credits can roll over month to month like rollover minutes on wireless plans. While AWS won't eliminate the micro-measurement of service usage, it could hide it behind a bulk line item and handle the messy usage statistics and billing adjustments internally.
• Given the high-profile spat between Oracle and AWS over the latter's use of Oracle databases internally and AWS' accelerated timeline for migration to Aurora, expect AWS to more aggressively push its database migration service as a way to both win more enterprise business and to stick it to a vocal critic and competitor.

Ernesto Marquez

I expect new features for detailed monitoring of serverless applications in the future of AWS. The cloud provider has launched many features to help customers with serverless application management and deployment, but detailed monitoring still lags behind. There is a real customer demand for this, too, and there are some AWS partners that already give AWS customers more visibility into their serverless applications.

I also expect some cool features in 2019 where existing services get smarter with the incorporation of machine learning and automated reasoning. Some recent examples include machine learning insights for QuickSight, predictive scaling for EC2 and S3 Intelligent-Tiering. Maybe smart scaling for Kinesis or smart allocation for Lambda is next.

Alan Earls

Microsoft Azure and IBM Cloud continue to paint a strong enterprise story for a large audience that already looks to them for answers. In response, AWS will likely take steps in 2019 to ensure its relevance in that space. Some of this could simply be window dressing as AWS tries to find ways to look and sound more enterprise-focused or to better explain its enterprise value proposition.

In addition, I think AWS will make visible efforts to enhance its offerings to be more robust, usable, coherent and enterprise-friendly. AWS will do this on its own or through the kind of partnership they have established with VMware, which produced Outposts.