Everything Amazon does these days is news, and for good reason -- they’re constantly pushing the envelope.
This year’s Amazon Web Services re:Invent convention in Las Vegas was the largest global conference focused on cloud computing. It attracted 40,000 engineers, vendors and IT experts and featured hundreds of diverse events.
Topics ranged from high-level strategy to in-depth technical talks along with numerous big announcements regarding new services and features.
Missed out on this year’s conference? Here are our top takeaways regarding what’s new with AWS.
EC2 M5 compute instances: These next-generation instances are based on a custom-built chip code, “Nitro,” which runs the hypervisor. M5 instances improve price and performance as compared to M4 instances by 14% and were designed to keep up with ever-increasing intensive workloads, particularly surrounding image and video processing and data compression.
- Bare metal instances, now available: Amazon EC2 Bare Metal instances allow applications to directly access servers’ processors and memory. This means you can deploy applications that rely on physical hardware directly on AWS infrastructure, and scale them up or down in mere minutes. According to Amazon: “These instances are ideal for workloads that require access to hardware feature sets or for applications that need to run in non-virtualized environments for licensing or support requirements.”
- Sumerian: Mixed reality is now becoming a more widely accessible technology. Sumerian is a tool set that allows developers to quickly build and host VR, AR, and 3D apps for a range of devices. You pay only for the storage you create, and Sumerian also comes with a library of pre-built 3D objects to make building characters and environments faster and easier.
- Aurora MultiMaster: This addition to Amazon’s Aurora relational database makes it possible to set up multiple read replicas and write replicas to scale both input and output across regions with millisecond response times. This is big news for high-end database administrators who need to scale write-heavy databases. And its clusters mean even higher availability and no risk of downtime.
- DynamoDB Global Tables: Global Tables relies on existing Amazon DynamoDB tables to serve as a scaling feature for NoSQL databases. The results? A fully managed, global, and multi-master database that allows for quick read and write performance for massively scaled applications. According to Amazon, it also “eliminates the difficult work of replicating data between regions and resolving update conflicts, enabling you to focus on your application’s business logic. In addition, Global Tables enables your applications to stay highly available even in the unlikely event of isolation or degradation of an entire region.”
- Simple Storage Service (S3) Select: This is the next evolution of databases, migrating from data warehouses to “data lakes.” Data warehouses are not agile and flexible enough to store and move the mass amounts of data companies accumulate and work with on a daily basis. S3 Select provides the ability to query S3 data objects with SQL commands. Because it retrieves only the data an application needs at that moment, it keeps costs low while greatly improving performance.
- Glacier Select: Glacier Select expands the power of S3 Select for AWS cold storage. It was built for use in highly regulated industries, such as healthcare, government, and financial services, and keeps companies in compliance by writing data directly to Glacier. Glacier Select is different than most cold storage options in that data can be easily located and accessed within minutes by simply entering an additional set of parameters to filter the information and bypass security.
Translate, Transcription and Comprehend: Remove language barriers and streamline communications with these three new AWS tools.
Amazon Translate relies on deep learning models and machine learning to provide more accurate and native-sounding translations than ever before. The neural machine is constantly learning to provide more accurate translations, can translate immense blocks of text, and can be embedded into existing applications easily through an API call.
Amazon Transcribe can save companies time and money on automated transcription services, and ensures a higher level of accuracy, thanks to its formatting and punctuation capabilities. It currently is available for English and Spanish speakers, though it will be expanded soon, and can be used for a wide range of use cases beyond transcribing interviews, like collecting customer support data.
Amazon Comprehend: is an important companion piece to this language-based machine learning trinity. It breaks down the various elements of languages to make the building blocks easier to understand, including language origins, key phrase extraction, and emotional analysis. Essentially, it can add context, a crucial but often missing piece for cross-lingual teams.
SageMaker: Aptly named, this end-to-end tool enables developers to build, teach, and deploy their own machine learning/AI models. While it’s somewhat a response to IBM’s Watson AI tool set, Watson is primarily used by data scientists, while SageMaker is geared toward general developers. SageMaker comes with various modules to make the process of collecting data, choosing algorithms, determining frameworks, and testing models a bit less complex.
Gluon: Deep learning is also getting simplified with Gluon, an AI and machine learning tools and frameworks library. Gluon comes with easy-to-understand, swappable blocks of code to build neural networks, and the flexible framework makes it easy to build, edit, and debug as you go, without any loss of training speed.
New P3 GPU-based instances: Powered by Nvidia’s Tesla Volta chip, the newly-released P3 instances on AWS’ EC2 cloud can put over 4,000 optimized cores at your service. The P3 instances can handle intensive machine learning, deep learning, and other workloads, and AWS claims they can also significantly reduce time spent on deep learning model training.
Alexa for Business: Your favorite virtual assistant (sorry, Siri!) has entered the workplace. Alexa for Business is a multi-purpose IoT appliance that can save you time by automatically setting up conference calls and webinars, order office supplies, open IT tickets, and more — all via vocal commands.
DeepLens: AWS presented a new hi-res camera with built-in AI and machine learning capabilities, retailing at $250. The camera was created with developers in mind, and most notably comes with image, object, and human recognition features. The DeepLens shoots 1080p video and photo, has a range of connectivity options and has the potential to streamline business operations through its trainable, vision-based AI capabilities.
Lambda@Edge: You now have the ability to embed Lambda functions in previously static-only services, such as S3 and CloudFront. Use Lambda functions to trigger actions at various points in the CloudFront request and response process.
Amazon is upping their game, which means your company can too -- as long as you have the expertise and resources to take advantage of the tools they’re providing.A more approachable option? Codero can architect, implement, and manage your AWS cloud to make sure it’s optimized for your company. That way you get the best of today’s hosting technology without the maintenance overhead.