AWS re:Invent 2016

In this session, customers learn more about the T2 instance type and the performance and cost savings it can bring to startups, SMBs, and enterprises. Customers will share best practices and tips for how they use T2 instances across workloads including development and test, production web servers, continuous integration and more.

Each of the four distinct compute services available from AWS bring unique advantages to your application's design: Amazon EC2, AWS Elastic Beanstalk, Amazon ECS, and AWS Lambda. However, your security responsibilities change with each of these services. For example, with Amazon EC2 and AWS Elastic Beanstalk, you need a plan to lock down the operating system, your applications, and your data. Containers running in Amazon ECS provides additional flexibility and mobility but also introduces new challenges as most security techniques were not designed with containers in mind. AWS Lambda requires a completely new approach to security design at the function level.

Apache Spark is well known across industries, use cases and businesses of all sizes for its speed and ease of use in sophisticated analysis of large datasets. In this session, learn from Ion Stoica who co-led the Apache Spark project at the AMPLab (UC Berkeley) and co-founder of Databricks, about some of the latest innovations in Spark 2.0, a new open source tool Earnest to choose the optimal cluster configuration for your job, and how and why Databricks choose EC2 to run Spark. We’ll also take a look at how Amazon EC2 and the latest enhancements enable Sparks as a data processing platform, along-with best practices and cost optimization techniques for using Spark with AWS.

Amazon EC2 allows you to bid for and run spare EC2 capacity, known as Spot instances, in a dynamically priced market. On average, customers save 80% to 90% compared to On Demand prices by using Spot instances. Achieving these savings has historically required time and effort to find the best deals while managing compute capacity as supply and demand fluctuate.

ProtectWise has been hailed as one of the top 10 coolest startups of 2016, disrupting the security industry by providing cutting edge cloud-based network threat detection, forensics and analytics for Fortune 2000 customers. By developing a scalable, stateless platform, they are able to quickly right-size Amazon EC2 instances to optimize performance and ingest gigabytes of network packet capture at a fraction of the cost. Learn how their cloud-first approach towards infrastructure and application development has enabled them to be more agile, grow faster, and save money. In this session, attendees gain practical best practices for how to build cloud-ready applications and see real-life examples of how this hot startup applied those to common workloads like Cassandra.

In six months, Verizon has built a best-in-class Augmented Reality and Virtual Reality (AR/VR) platform that streams HD video and game experiences using Amazon EC2 GPU Accelerated instances and CloudFront. Verizon will share their reference architecture and configuration best practices that enabled them to develop a massively scalable VR architecture that scales to support for 100K simultaneous HD video streams to customers around the globe.

AWS GPU computing capabilities have allowed hot start-ups like Zerolight to revolutionize the car buying experience for clients like Audi, by providing advanced 3D car rendering simulations for their customers interested in personalizing cars in real-time at dealer showrooms and online. Zerolight chose to build their company from the ground up on AWS because of GPU-compute capabilities, scale, elasticity, and reliability, allowing customers to perform complex renderings quickly and in real-time. In this session, learn how Zerolight has implemented a scalable, cost-effective, highly responsive 3D rendering platform using Amazon EC2 G2 instances.

Algorithmia is a startup with a mission to make state of the art machine learning discoverable by everyone - they offer the largest algorithm marketplace in the world, with over 2500 algorithms supporting tens of thousands of application developers. Algorithma is the first company to make deep learning, one of the most conceptually difficult areas of computing, accessible to any company via microservices. In this session, you learn how this startup has selected and optimized Amazon EC2 instances for various algorithms (including the latest generation of GPU optimized instances), to create a flexible and scalable platform. They also share their architecture and best practices for getting any computationally-intensive application started quickly.

Many customers are using Amazon EC2 instances to run applications with high performance networking requirements. In this session, we provide an overview of Amazon EC2 network performance features (enhanced networking, ENA, placement groups, etc.), and discuss how we are innovating on behalf of our customers to improve networking performance in a scalable and cost-efficient manner. We share best practices and performance tips for getting the best networking performance out of your Amazon EC2 instances.

FINRA is a leader in the Financial Services industry who sought to move toward real-time data insights of billions of time-ordered market events by migrating from SQL batch processes on-prem, to Apache Spark in the cloud. By using Apache Spark on Amazon EMR, FINRA can now test on realistic data from market downturns, enhancing their ability to provide investor protection and promote market integrity (FINRA enacts rules and provides guidance that securities exchanges & brokers must follow). By using AWS Spot instances, FINRA has saved up to 50% from its on premises solution, increased elasticity/scalability, and accelerated reprocessing requests (from months to days). Learn best practices on how FINRA moves toward real-time data analytics with Spark and AWS, while managing production workloads in parallel, increasing performance and IT efficiency, reducing cost, and modernizing and scaling their infrastructure to prepare for real-time processing in the future.