AWS re:Invent announcements 2020
Things are a little different this year at AWS re:Invent, but one thing remained unchanged: Andy Jassy’s AWS 2020 re:Invent keynote was packed with exciting announcements! There are over 500,000 people worldwide registered for this year’s virtual conference, and every Toad is joining in for the fun. The products and features that Andy Jassy announced in his keynote will—once again—change the face of cloud, and we can’t wait to put them to work for our clients.
If you missed the keynote or need more details on any of the products announced, we’ve got you covered! This recap of Andy Jassy’s AWS re:Invent 2020 keynote hits all the critical points, and you can follow the links if you want to dig in—or get in touch with us for more support. And if you registered for the conference, you can also watch the whole thing here.
What does it take to reinvent?
Andy started out by reflecting on the question of what it takes to reinvent, and offered the following insights:
- Leadership will to invent and reinvent
- Acknowledge that you can’t fight gravity
- Talent that’s hungry to invent
- Solve real customer problems
- Don’t complexify
- Use the platform with the most capabilities and broadest set of tools
- Pull it all together with aggressive top-down goals
We couldn’t agree more—transformation requires both a new mindset and the right tools at the right time. Speaking of the right tools...
After the inspiring intro, Andy took us through an exciting array of new and enhanced products and services. Below is a roundup of the latest releases from AWS (and you can keep up with even more rollouts on the AWS re:Invent 2020 blog).
- There are powerful new Amazon EC2 instance types: compute optimization (C6gn), GPUs (G4dn), dense hard disk drive storage (D3 and D3en), low-latency networking (M5zn), and higher Elastic Block Store (EBS) bandwidth and IOPS (R5b).
- EC2 Mac instances are available for the first time, a huge opportunity for Apple apps (though limited to test and development at present).
- AWS Graviton2 processors will power Amazon EC2 T4g, M6g, C6g, and R6g instances, and provide up to 40% better price performance compared to the current generation
- AWS Compute Optimizer has new support for EBS volume recommendations.
- AWS will integrate Habana Labs' Gaudi accelerators into new EC2 instances for deep learning, with a projected 40% improvement with price performance.
- AWS will launch a custom chip, AWS Trainium, designed for machine learning training, that will support all major frameworks and the AWS Neuron SDK.
- Developers can now use Amazon Elastic Container Registry (ECR) Public (an extension of the present-day ECR service) to share container software anywhere in the world. The companion Amazon ECR Public Gallery allows users to browse container images and other details.
- AWS now provides an open source Kubernetes distribution of Amazon EKS Distro.
- Amazon Elastic Container Service (Amazon ECS) Anywhere and Amazon EKS Anywhere—previously available on an AWS Outpost, but at a relatively high cost and specific form factor requirements—will allow customers to run ECS/EKS in their own data centers, alongside their on-premises apps and infrastructure.
- The billing increment for AWS Lambda has been reduced to 1 millisecond (previously 100 milliseconds). This has the potential to mean big savings for applications with large volumes of short-duration functions! (Rates for provisioned concurrency and requests remain the same.)
- AWS Lambda functions will support 10 GB—three times the current memory—and can now be provisioned as container images.
- Amazon CloudWatch Lambda Insights offers monitoring capabilities geared toward functions, plus a health dashboard to track unwanted changes caused by new function versions.
- AWS Proton provides a PaaS-like environment for developers that allows the ops team to retain control of infrastructure. Infrastructure teams can create templates to be used across environments to automate the full lifecycle for containerized and serverless applications.
- AWS SaaS Boost is an open-source reference environment for independent software vendors to build apps on top of AWS.
AWS Data and data sources
- The EBS gp3 Volume will separate performance provisioning from storage capacity for general-purpose SSD volumes in EBA, which may save customers up to 20%.
- New io2 volumes—the next gen of Amazon EBS volumes, which is optimized for large relational databases—will quadruple throughput, IOPS, and capacity. AWS also added tiered pricing for IOPS for customers with more than 32,000 IOPS per month.
- Amazon Aurora Serverless v2 will slash capacity scaling time to fractions of a second (previously 5-50 seconds) and now includes some standard Amazon Aurora features.
- Amazon S3 is now strongly consistent, an impressive accomplishment given the maturity and scale of S3. Other S3 updates include two-way replication of object metadata between buckets and the option to replicate S3 buckets across multiple destinations—within a region or to another one.
- Babelfish for Amazon Aurora will translate commands from Microsoft SQL Server to Amazon Aurora PostgreSQL, creating massive efficiencies for converting legacy SQL Server databases to AWS native databases.
- AWS Glue Elastic Views will create materialized views of data combined from disparate sources through SQL queries.
- Advanced Query Accelerator for Amazon Redshift will deliver a distributed, high-speed cache for faster queries.
AWS Machine learning
- Amazon SageMaker was updated with tailored capabilities for CI/CD, feature repositories, and data prep.
- Amazon DevOps Guru, a machine-learning-backed service that searches for abnormal operational patterns, identifies potential issues and causes and can suggest solutions.
- Amazon QuickSight Q will add natural language capabilities to the business intelligence service.
- Amazon Connect has new tools! Amazon Connect Wisdom uses machine learning to reduce the time it takes customer service agents to find answers. Amazon Connect Customer Profiles aggregates customer data for agents to help them deliver personalized customer service. Amazon Contact Lens now offers real-time call analytics capabilities, enabling businesses to detect issues during live calls and resolve issues faster. Amazon Connect Tasks helps agents and supervisors optimize agent time. Finally, Amazon Connect Voice ID provides real-time caller authentication.
- Amazon Monitron puts detection and preventative maintenance functions in an end-to-end predictive machine learning system.
- Amazon Lookout, intended for industrial settings, has two new offerings. Lookout for Vision performs real-time analysis of product defects. Lookout for Equipment enables predictive maintenance by analyzing data from industrial machinery and using machine learning to find and predict abnormalities.
- AWS Panorama is a machine learning device and SDK that integrates with on-premises camera systems to monitor workplace conditions; AWS emphasizes its potential to address manufacturing quality and employee safety.
Leveling the playing field
- AWS Outposts will roll out two new form factors (1U and 2U servers) that allow users to run AWS on-premises with much smaller appliances. They are optimized for non-data-center settings, like hospitals and manufacturing operations.
- AWS Local Zones (which provide a subset of Amazon cloud services in major population centers for ultra-low latency applications) were added in Boston, Houston and Miami. AWS plans to launch another dozen in 2021.
- An AWS Wavelength Zone was added in Las Vegas using 5G technology.
How will you reinvent?
Whew, that is a lot of leading-edge innovation to digest! Luckily, here at Metal Toad we eat innovation for breakfast.
As an award-winning AWS Consulting Partner, our mission is partnering with industry-leading companies to transform their outcomes through the power of the cloud. If you want to explore how the latest AWS developments can drive success for your business, drop us a line!