My thoughts on AWS re:Invent 2016 keynote by Andy Jassy
AWS keeps re-inventing the Cloud industry
This year (Nov-27 to Dec-2), like every year, AWS hosted the AWS re:Invent event in Las Vegas. This is a 5 days event where AWS presents their current state and a series of new services, improvements and better ways of doing things. It’s an event where customers are also invited to share their success stories with AWS.
This year there were a couple of keynotes that are worth watching. In this blog I’m going to talk about Andy Jassy’s one whereas in the next blog I’ll be talking of Werner Vogel’s.
Andy Jassy’s keynote in a nutshell
Andy Jassy, AWS CEO, talked about AWS giving customers superpowers. Specifically he talked of:
- X-Ray vision
- Shape Shifting
Leaving the super-hero themes on the side, this is what I found valuable.
- AWS has introduced a new set of configurations for the T2, R, I, C and GPU families. The new instance types offer more memory, more processors and more GPU power.
- There is a new service, AWS Lightsail, which allows customers to spin up new servers in three simple steps. Its cost starts from $5/month
- There is a new type of instance, F1, designed for high computing power
Andy introduced a whole series of new services related to Analytics, Machine Learning and AI.
- Amazon Athena. It’s a fully managed service that allows SQL queries on S3. It’s fast, simple and effective. It doesn’t require data to be stored in any particular format.
- Amazon Rekognition. It’s an image recognition and analysis fully managed service, powered by Deep Learning. By uploading images, the engine can extract a series of metadata, e.g. face recognition, sentiments, situations and so on. This allows to store information about images and their context in a database for later retrieval.
- Amazon Polly. It’s a text-to-speech fully managed service powered by Deep Learning. It performs a number of optimisations during the translation process, e.g. transforms the word “WA” in Washington in the right context.
- Amazon LEX. Works inside Alexa. It’s a natural language understanding and automatic speech recognition fully managed service.
These services are integrated between each other and in the AWS ecosystem, including Lambda. This provides developers with the ability to create services fuelled by AI capabilities. It will be possible for example to prime AWS AI with a list of sentences to recognise. The engine will assign semantic meaning to them and expand them to a higher number (~ 50) of different ways of saying the same thing. Then, users will be able to talk into a microphone and ask for services, e.g. “Book me a flight to London”. The possibilities of having access to AWS AI services and APIs are potentially endless. Watch this space as developers around the world will start creating amazing services.
Here Andy talked of few new services:
- AWS Greengrass. This is the ability to embed Lambda computing in IoT devices. Manufacturers will need to embed the AWS Greengrass runtime in the devices at build time but this will allow the ability to “extend” these devices capabilities.
- AWS Snowball Edge. This is a new device to transfer 100 Terabytes of data to AWS. It also offers computing capabilities on the device.
- AWS Snowmobile. This is a truck going around with a container to store and send AWS 100 Petabytes with each load. With high speed networks (e.g. 10 Gbit/sec), it would take roughly 25 years to transfer to AWS an Exabyte of data. With AWS Snowbile, it’d take approximately 6 months.
- Perhaps, though, the most impactful announcement in this section was the integration of VMWare services into AWS. This breaks a gap that historically existed between internal and external cloud. Now companies will be able to seamlessly extend VMWare-driven infrastructure to AWS. This will broad the adoption of hybrid models and will ease migration to AWS from customers who used to be put off the migration to AWS as they didn’t want to loose their VMWare capabilities.
AWS is an Enterprise Grade ready public Cloud offering
Andy’s keynote highlighted a number of factors. If anyone doubted AWS maturity as an enterprise cloud offering, by watching Andy’s keynote it’s clear that a number of large enterprises (including McDonald and Finra and many more) not only are operating on AWS with a number of advantages but now their strategy is to go “All-in AWS”. I was particularly impressed by Finra CIO’s presentation. They process trillions of records every day, with zero operational concern. Not only, they’ve made some of their APIs open source. This has been a recurring theme at this year re:Invent. Companies who have created successful implementations are sharing their API as open source projects. Even AWS has created Blox, a collection of open source projects for container management and orchestration.
AWS is the best public Cloud offering provider. They are the fastest IT growing company in the world and they keep reinventing themselves to offer their customers new and improved ways to operate in the Cloud.
Their mantra is to listen to their customers, adapt their strategy and product offerings based on that and in doing so, keep ahead of the competition.
This year AWS has all been around a DevOps operating model, automating everything from code commit to production, improving and innovating monitoring, logging and introspection in what happens in the services, containers, microservices, improvements to their RDS Aurora offering (which now supports PostgreSQL), new Analytics, Machine Learning and AI capabilities, new computing capabilities, serverless architecture as the main AWS primitive, partnership with key clients to help them move and succeed in the Cloud and much more.
This year alone AWS has released 1,000 new capabilities. Listening to your customers and shipping valuable software in their hands is the hallmark of every high performing organisation and AWS leads the way as one of the most innovative companies on the Planet.