Amazon’s AWS Re:Invent 2019 conference in Las Vegas welcomed more than 60,000 attendees, making it the largest Re:Invent over the years.
Last week, AWS revealed a product direction that re-emphasized and focused on prioritizing customer feedback. The common themes discussed revolved around ease of use at relentlessly declining price points, higher capability, and increased output. Amazon announced several new capabilities and product launches across its expansive product line.
Amazon launched many new features and product updates in the Amazon AWS Re:Invent 2019:
AWS Local Zone
AWS announced the opening of an AWS Local Zone in Los Angeles. AWS Local Zones are an innovative type of infrastructure deployment that places storage, computing, database management, and other selected services close to customers. This will provide developers in LA the ability to deploy applications that require single-digit millisecond latencies to end-users in LA.
The cloud giant unveiled nine new Amazon Elastic Compute Cloud (EC2) innovations. AWS added to its industry-leading networking and computing innovations with new Arm-based instances (M6g, C6g, R6g) powered by AWS-designed processors in Graviton2, machine learning inference instances (Inf1) powered by AWS-designed chips.
AWS announced the availability of AWS Outposts, which are entirely managed and configured with storage racks built with AWS-designed hardware. This will allow customers to compute and store on-premises while connecting to AWS’s services in the cloud. AWS Outposts bring native AWS infrastructure, services, and operating models to virtually any data center, on-premises facility, or co-location space.
AWS announced AWS Wavelength, which provides developers the power to build applications that serve end-users with single-digit millisecond latencies over the 5G network.
Re:Invent this year focused on democratizing machine learning (ML) to make it accessible to a broader set of business, creator personas, and developers. AWS has pushed to make machine learning more democratic and widely accessible with the updates this year.
Some significant Machine Learning focused offerings that were launched
Read More: Amazon’s new interest for AI in Healthcare
Inf1 instances and Inferentia chips
The company launched its custom-built Inferentia chips that make inferencing cheaper and faster. The new Inf1 instances available on EC2 integrate with Tensorflow, Pytorch, and MXNet. AWS claims three times the output at two-fifths the cost with this new technology.
Amazon Aurora ML integration
AWS unveiled its capability to integrate machine-learning predictions directly from SageMaker to comprehend into Aurora databases using SQL. This is simplified using direct calls that do not pass the application layer, ensuring low-latency with real-time use cases like fraud detection or product recommendations.
Alexa Voice Service (AVS) and IoT Core Integration
This integration will reduce the cost of building Alexa Voice into devices across a wide variety of categories, particularly in the resource-lean devices.
Model prototyping in Java using the Deep Java Library (DJL)
While Python is the chosen language for all ML devs, AWS finally has acknowledged the popularity of Java by announcing the DJL. DJL is an open-source library and API which is designed for prototyping and developing deep learning models in Java. The DJL works atop Pytorch and Apache MXNet.
Read More: Amazon Designs Robust Data Center Chip
SageMaker inferencing in Amazon QuickSight
Amazon’s business analytics and visualization services fall under QuickSight. AWS has now enabled ML predictions in Quicksight. Users can connect to various data sources and select prebuilt, custom, or packaged models and pipe visualizations into QuickSight.
SageMaker Studio IDE
SageMaker Studio IDE remained the most significant announcement throughout the event. AWS continues to extend SageMaker towards the company vision to develop a standard machine-learning environment to rule all. This comes with a slew of new capabilities and features that intend to transform the platform into a full-fledged web-based IDE for end-to-end ML workflows. AWS announced several unique features of the SageMaker Studio to help reduce the heavy lifting commonly associated with ML. These updates include SageMaker Notebooks, SageMaker Experiments, SageMaker Autopilot, SageMaker Model Monitor, and SageMaker Debugger.