AWS doubles down on edge computing, and other important re:Invent takeaways
Each December, AWS re:Invent draws tens of thousands of technologists from across the industry to network and get a peek at the newest AWS products and services. SHI was happy to once again send a strong contingent to Las Vegas to attend sessions, hear from customers about real-world implementations, and gather important information and best practices to guide our support of your cloud deployments in 2020.
This year we also had the good fortune of being named CloudHealth’s North American Partner of the year for the third consecutive year!
If you were unable to attend AWS re:Invent yourself, Amazon has made the keynote addresses and product announcements available here. But we’ve also highlighted the new offerings we think will have the most dramatic impact on AWS customers moving forward. Read on!
AWS moves further to the edge
There was a big push to drive more value and intelligence at the edge this year, as evidenced by AWS’s continued expansion of its edge and IoT portfolios. With new services such as Wavelength and Local Zones, AWS is giving customers the tools required to operate in this emerging environment.
WaveLength provides developers a platform for delivering ultra-low latency applications to 5G devices at the edge. AWS Local Zones is a new type of AWS infrastructure designed to run workloads that require single-digit millisecond latency. It’s currently only available in Los Angeles, with more regions to come.
AWS also announced the general availability of Outposts, which it first revealed at AWS re:Invent 2018. This service extends AWS services and APIs to customer locations.
Other thoughts on AWS re:Invent
From the desk of Tony Vinayak, Senior Director Professional Services
The most fascinating announcement for me was Amazon Braket for quantum computing. While quantum computing is still in its infancy, Braket provides an interesting opportunity to start dirtying your hands via the simulated environment. This is a very exciting evolution in the quantum computing space that will speed development. If you’ve been reading about qubit lately, it’s nothing short of revolutionary.
Another session that drew large crowds was the one on the relatively new AWS Cloud Development Kit (CDK), which enables a software development framework for defining cloud infrastructure in code and provisioning it through CloudFormation. Here at SHI, we recently adopted CDK for developing a higher-level abstraction layer for infrastructure automation. (Check out our GitHub repo.)
One perspective I got while talking to AWS Professional Services staff members was the fact that AWS has also started focusing on refining usability of existing services while easing up a bit on introducing new services. That’s a relief! Customers are also looking for more prescriptive advice on using those services, rather than having to wade through them all and figure out how to piece together various services and options.
From the desk of Daniel Newman, Principal Consultant
A common theme at this year’s re:Invent was the incorporation of artificial intelligence (AI) and machine learning (ML) into many existing products and new services.
Amazon CodeGuru is one that stands out. It’s a machine learning service that provides automated code reviews and performance recommendations. Today it supports Java applications, but it will include more in the future.
AWS also released Amazon Sagemaker Autopilot, a new service to help automatically create machine learning models. The ultimate goal of these services is to provide the business value of ML with minimal investment.
Two security announcements also warrant mentioning:
- Amazon Detective: Root cause analysis of potential security issues or suspicious activity (Correlation and Analysis of multiple AWS service logs/events)
- Amazon Fraud Detector: An ML-enabled service that takes lessons learned from Amazon’s years-long online shopping presence and provides them to customers as a managed service to identify potentially fraudulent online activities such as online payment fraud and the creation of fake accounts
On the managed service front, AWS announced a new managed Apache Cassandra Service, rounding out a good mix of existing managed DB offerings.
From the desk of Vincent Montalbano, Cloud Solution Consultant
Each year I look forward to Amazon CTO Werner Vogels’ keynote presentation. Vogels started by explaining how Amazon “has been focusing on things that never change” – security, performance, scale, reliability, cost efficiency, and operational excellence. Amazon is dedicated to continuous improvements to these foundational components under the assumption that these parameters will always benefit AWS customers. As a solution architect, I could not agree more.
These two announcements from the keynote illustrate that vision:
- AWS Nitro: The AWS Nitro System is the underlying virtualization platform for the AWS EC2 instances. AWS Nitro improves the technical underpinnings of the traditional hypervisor by separating the traditional CPU, storage, and networking functions into independent Nitro cards – a family of cards that offload and accelerate IO for functions, ultimately increasing overall system performance. The Nitro Security Chip offloads virtualization and security functions to dedicated hardware and software, minimizing attack surface and enhancing security.
- Amazon Builders’ Library: The Amazon Builders’ Library is a collection of documents that provide insight into how Amazon develops, architects, releases, and operates technology. The library’s content is written by Amazon’s senior technical leaders and engineers and covers topics across architecture, software delivery, and operations. If you ever asked yourself how Amazon does something, here is your answer.
From the desk of William Kerr, Senior Developer:
When a job can’t run in a lambda function due to size or time constraints, the next easiest way to use compute on demand is Fargate. It’s nice to see that AWS has expanded Fargate to run in a serverless EKS cluster.
SageMaker Autopilot is a nice stepping stone that should help anyone new to ML get models trained and deployed with ease. In the past, it was a bit of an inconvenience not having an automatic tool available.
From the desk of Todd Wolff, AWS Business Development Lead:
The two notable announcements for me were:
- AWS Service Ready Program: AWS Service Ready identifies and validates products from APN Technology Partners that integrate with specific AWS services and recommends these products to customers. That means AWS customers can spend less time evaluating new tools and more time scaling their use of products that are integrated with AWS services. Products in AWS Service Ready have been reviewed by AWS Partner SAs for high availability and proper architecture and follow AWS best practices for developing integrated products.
- AWS Retail Competency: AWS Retail Competency Partners provide innovative technology offerings that accelerate retailers’ modernization and innovation journey across all areas in the enterprise. These include Customer Engagement; Supply Chain and Distribution; Physical, Digital, and Virtual Store; Advanced Retail Data Science; and Core Retail Business Applications. AWS Retail Competency Consulting Partners offer strategy and deployment services to retailers to help accelerate their digital transformation.
Follow the SHI Blog for continuing coverage of these and other AWS offerings as they are rolled out and implemented in more real customer environments.