AWS re:Invent Keynote highlights


While this is the 7th AWS re:Invent, this is my first time at re:Invent and I’m pretty excited about learning how AWS CEO – Andy Jessy takes on the keynote, and as I normally do when I attend a conference, I would like to provide you with my AWS re:Invent Keynote highlights.

This year the conference will have 2400 sessions and over 50000 people in attendance. Andy also took us through the list of customers that go from small startups to large enterprises without forgetting the public sector.

As Andy continued talking about numbers he shared that AWS has tremendous growth during the last 12 months and has become the clear market leader on this space with $27 billion annual revenue with a 46% year over year growth.

The clear leadership of AWS on this space is also highlighted by companies like Gartner, giving AWS a 51% market share.

Moving to the next topic, Andy mentioned this conference is first and foremost a place to learn and because of that builders are a key part of the conference, so he introduces the re:Invent band to tell us what they want

“I want it all and I want it now” by Queen is what they are signing to, very nicely timed based on the success of the recent movie “Bohemian Rhapsody”.

Why is this important? AWS platform provides builders over 100 different services to everyone and every need, from databases (11 different types) to compute to storage to containers to the very important (for me at least) End User Computing services. A list of those services can be found here.

It is time for announcements and Andy spent quite a bit of time on the Storage services and presented the most exciting news thus far on the keynote.

A big announcement is Glacier Deep Archive, a service to replace tape. This is set to launch in 2019 with a cost of as little as $1 per Terabyte

Now moving to file sharing, AWS has have EFS for awhile now but it has been focus on NFS and Linux file system. So Amazon FSx for Windows File Server was announced to provide native Windows File services. This is a great addition and it is HIPPA, PCI compliance to name a couple. This is a very welcome addition for enterprise.

For compute intensive workloads that have file services requirements, Amazon is also announcing Amason FSx for Luster. The solution is capable of be started and stopped on demand so It can be used when in need of processing and then removed when not.

The second “sentiment” for the a new type of builder is: “I have been waiting, waiting and I want you”

AWS Control Tower is a prescriptive and secure way to setup environments. This is for the builder that has a little less experience setting up Landing Zones. This is will take best practices and translating them to the different services and rules in AWS.

Continue with the security team AWS Security Hub. A GUI to easy setup and monitor the environment and also easy to integrate with third party security solutions,.

Analytics is still a trend on the IT industry and data lakes are key to the success of the strategy but it is not easy to get it right

And then once you got it right, how do you secure it? it is as complex as setting the data lake itself. The announcement is AWS Lake Formation. Setup and configure in days not months is the promise for this service.

 Moving into databases, Andy suggested that enterprises want database freedom, that Oracle and Microsoft are restricting companies from moving faster. Amazon Aurora helps the customer to use engines like MySQL and PostgreSQL at scale and enterprise grade. This services has added over 35 new features since its launching and is one of the faster growing services for AWS.

Relational databases has been the default implementation for most applicaitions

But business requirements and applications requirements have changed, so AWS has adapted to those requirements. “The right tool for the right job”

One of the most dificult capacity  Dynamo DB Read/Wriet Capacity On Demand.

Tracking and analyzing change over time series type of data

Amazon Timestream is the newest database service that is a purposely build DB for time-series needs

Moving to one of the most interesting technologies of recent times, Blockchain. AWS has been talking to customer and technologist to understand well what needed to be done and what needed to be solved.

For these tasks AWS is introducing Amazon Quantum Ledger DB (QLDB) and Amazon Managed Blockchain

With this last announcement around databases the AWS landscape has change a bit from our first DB related slide.

Machine Learning “little less conversation and more action” is the tune the house band played to introduce this topic. According to Andy, more machine learning is happening on AWS than anywhere else. So how does AWS look at ML?

Scaling TensorFlow has been difficult and it is costing money to the customers so the AWS took on that challenge to optimize TensorFlow and empower the customers as well as saving them money.

As we continue to review the challenges of ML, inference inefficiency is one of them specifically because of the cost.

To resolve that Andy announced Amazon Elastic Inference. this a technology to enable ANY AWS instance with GPU acceleration at a fraction of the cost that was previously available. In my opinion this will accelerate even more companies and their ML strategies

Amazon Inferencia a new custom build chip by Amazon to bring the cost of ML even further down. This will integrate with Elastic Inference and will also be available for all Instance types.

We saw a great demo from Formula-1 on how they are leveraging AWS HPC and ML services to improve the sport and engage more fans. It was impressive over 120 sensors per car and generating 1.1 million data points per second.

As we continued the car theme, Andy took us over the challenges of labeling images one car automation, to help with this issue Andy introduced Amazon SageMaker Ground Truth. There were more SageMaker announcements on the this topic like AWS Marketplace for machine learning for people to publish algorithms.

Because there are cool things and then this, Amazon introduced AWS DeepRacer and 1/8th scale autonomous car for $399 and if you preorder it will be $249. Amazon has also created a race league for people around the world to compete on. More information here

Another new service is Amazon Textract to improve the good old OCR type of function and extract data from documents to actually take actions on them and the most important thing is that the builder doesn’t need ML experience.

This was a really announcements packed part of the keynote and the image below summarize it a bit.

The next sentiment is by the Clash “Should I stay or Should I go”. This is oriented towards the questions on most of IT departments out there when it comes to do I transition to the cloud or do I stay with the traditional on-premises infrastructure.

And in order to do this customer would like to continue to use the same tools they are familiar with, in most of the cases this is VMware. So we have VMware CEO Pat Gelsinger giving us an update on VMware Cloud on AWS (VMC).

The evolution to this service is to bring AWS to on-premises, for this AWS is announcing AWS Outposts. AWS hardware on-premises with a common set of APIs for public and private clouds. With this VMware is also going to offer their services on AWS Outposts, enabling Networking synergies and VMC. More information here

To wrap up the keynote, Andy comments on what is important for AWS and that should be important for everyone, to maintain a successful company don’t only look at what the competitors are doing but work with your customers, learn from them, pay attention to them and create what is important for them at an accelerated rate.

Thank you for reading. Leave your comments below.

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>