AWS re:Invent 2020 is just around the corner. And while this year's virtual format will create a different experience than previous in-person shows, attendees can still expect the usual flood of product updates, new services and other announcements from the cloud giant.
This year's conference will be a three-week virtual event, spanning from Nov. 30 to Dec. 18. Editor's note: After publication, AWS extended the dates for re:Invent, adding additional breakout sessions from Jan. 12 to Jan. 14, 2021.
Each year, AWS showcases particular technologies and developments at re:Invent. Highlights from last year include the release of AWS Outposts, new machine learning tools and an early foray into quantum computing.
So which technologies and topics will AWS focus on at re:Invent 2020? We asked SearchCloudComputing's expert contributors what they expect to see. Check out their predictions to prepare for the upcoming conference.
Kurt Marko, consultant
The 2020 re:Invent conference will be unlike any other in moving entirely online, but the experience will reshape re:Invent in future years. In retrospect, the 2019 version will serve as the high mark for in-person attendance. In the coming years, fewer people will want to endure the expense, inconvenience and time wasted schlepping backpacks on shuttles and through convention centers just to hear presentations that are often watched on video screens. Going from an $1,800 registration to free is a significant incentive for attendees to forego the in-person experience.
As 2020 progressed, companies learned from the mistakes of early online events and delivered well-produced keynotes and sessions with few technical glitches. I expect the production quality, session organization and presentation to be the best yet and set a standard for others in 2021. As usual, AWS will announce many new products, service updates and price cuts. Having a well-organized online site will simplify the process of finding and saving information and scheduling events.
On a technical level, look for AWS to emphasize its homegrown hardware, including Graviton2 ARM CPU, Inferentia deep learning inference chips and the Nitro system, as well as software like its Aurora database service, SageMaker for AI development and deployment, Lambda serverless functions and EventBridge for SaaS integration. Expect updates for Inferentia and Graviton3, although these services won't be available until mid-2021. Also, look for new Nitro cards that offload network functions similar to what NVIDIA does with its BlueField-2 data processing unit.
Expect discussions about serverless, in the broader sense of services that don't require pre-configuration or dedicated resources -- e.g., instances and databases -- and how these are used to build cloud-native applications that are more cost-efficient, scalable and adaptable. Pay attention to how AWS positions its pre-built, application-layer services versus how it sees containers and their role within application design. The positioning is important, since Google uses containers as a badge of openness and portability, in contrast to native cloud services that lock customers into a particular vendor and implementation.
George Lawton, IT writer
This year at re:Invent, AWS will embrace multi-cloud -- sort of. Amazon's long-standing lead in the cloud market afforded it the luxury to move slowly toward hybrid cloud and avoid multi-cloud. But as other cloud providers catch up, Amazon will be driven to adopt a stronger multi-cloud presence to maintain its lead -- particularly as larger enterprises shift more lucrative workloads to the cloud.
At re:Invent 2020, AWS will take a stronger position to enable integration across other cloud platforms. It's already allowing its partners to use the term multi-cloud in their messaging, after establishing rules against it at last year's re:Invent. Amazon will strengthen support for orchestrating workloads across multiple cloud services, which will allow it to maintain its edge as more enterprises adopt multi-cloud. I also expect to see improvements to CloudFormation for deployment and AWS Config for remote management.
AWS will also dive into digital twin technology. There are currently no big digital twin vendors. This is a capability that can be added into product lifecycle management, supply chain management and other types of tools. Amazon is in a position to provide a way to consolidate digital twin capabilities related to capturing IoT data, modeling the real world and providing different views of digital twins across stakeholders. AWS already started to do this; it just has not used the term -- yet.
AWS IoT Things Graph provides basic digital twin capability. This year, the company will either extend AWS IoT Things Graph to make it easier to develop digital twins or announce an explicit set of tools for modeling and managing digital twin capabilities. There will be new capabilities to better organize IoT related data streams, create and customize models that are updated by this data and present versioned views of this data via other applications. This will eventually support the development of a digital twin marketplace.
Brian Kirsch, IT architect
While AWS will introduce some core new product features, Kubernetes and the pandemic will take center stage at re:Invent this year. AWS will show what it is doing to help address both of those needs.
We will continue to see AWS push for containers and Kubernetes, similar to most vendors in the IT industry. While this is critical to those who are able to use and embrace container services, this focus will fall flat for a larger number of clients who don't.
There will be a strong focus on the rapid adoption of cloud computing, which was accelerated by the pandemic. I expect AWS will showcase those migration efforts and ease of transition to its services. Amazon will use this as a key point to remove some of the fear that companies might have about cloud service adoption. This will be the launching point to enhance existing migration services and launch new ones.
AWS will highlight how companies that did a quick migration and sudden scale-up saved a lot of money. Amazon will present cost-savings data to demonstrate why the leap to the cloud out of necessity won't need to be undone -- even after the pandemic ends. This will be the opportunity to introduce more features that impact price, remote ability and management. I expect to see several enhancements and releases tied to this.
Chris Tozzi, analyst
Amazon will likely announce updates to Outposts, its hybrid cloud platform -- or, failing that, will announce other hybrid initiatives. AWS has fallen behind Microsoft Azure and Google Cloud Platform when it comes to hybrid cloud, and there has been little momentum surrounding Outposts since the platform debuted. Enhancing Amazon's hybrid offerings could also help the company remain competitive in the face of the cloud repatriation trend, by giving enterprises an easier way to keep using AWS even if they move some workloads back to private data centers.
I also expect some announcements about enhancements to Amazon Elastic Kubernetes Service and Amazon Elastic Container Service. Likely nothing revolutionary, but Amazon continues to face pressure to innovate around Kubernetes and containers. It would not surprise me if Amazon released new management or security tooling for those platforms.
Lastly, I imagine there will be the usual announcements about cost-cutting for some services, especially in light of the budgetary pressures that Amazon's customers are facing due to the pandemic and economic turmoil.
Dan Sullivan, consultant
Data and machine learning will again get a lot of attention at re:Invent. AWS has expanded its support for machine learning with SageMaker services. The next logical step is to improve support for data engineering by building on existing services. AWS Glue is well suited for data warehousing ETL and AWS could enhance it to provide more support for data exploration. Deep integration of Data Catalog and SageMaker will help with the early stages of identifying data sources and assessing their quality.
The complexity of data management in the cloud is a barrier to building quality machine learning models. I expect to see efforts to support DataOps, particularly around automated data preparation and lifecycle policy management.
Ernesto Marquez, cloud consultant
I expect announcements regarding more built-in integrations between third-party SaaS products and Amazon cloud services such as AppFlow and EventBridge, as well as additional functionality once external data arrives in AWS.
AWS released several updates to its AI offerings this year, such as Augmented AI and DeepComposer. So, I expect this trend to continue with more relevant AI announcements at re:Invent 2020.
I also expect AWS to expand its Savings Plans coverage to more Amazon services that consume compute, such as Amazon Relational Database Service, Elastic MapReduce, Redshift, DynamoDB Accelerator, ElastiCache and Elasticsearch.