Big Announcements at AWS Re:Invent 2018
Big Announcements coming from AWS Re:Invent 2018. Check out the top 4 I think are pretty interesting. Big announcements in Machine Learning, GPU’s, New EC2 Instances, and IoT.
- 45% Cheaper RISC — EC2 Instances for Containerized Microservices
There is a new A1 EC2 instance running on Arm-based processors from Advanced RISC Machines (ARM). These are RISC chips or Reduced Instructions Set Computing processors. RISC processors provide faster processing of data while reducing power demands dramatically. Less power = less cost. RISC processors are everywhere and are found in smartphones, tablets, and wearables. Microsoft even offers a RISC based surface computer.
Careful, not all workloads are ready for these EC2 Instances.
2. New P3dn Instances could reduce Machine Learning training time to less than an hour on very large datasets.
Machine learning on very large datasets is an exploding requirement in the data science community. The new P3dn instances promise to boost performance of model generation with data located in Amazon S3 or EFS via a new offering: 8 Nvidia Tesla GPUs, 96 Intel Xeon Skylake virtual CPUs, and 1.8TB of NVMe-based SSD. The speed improvement is significant up to 100Gbps.
3. Need high performance computing, AWS might have you covered.
Need a supercomputer, come to AWS! The High Performance Computing historically has been a custom built machine. AWS has a new Elastic Fabric Adaptor that allows AWS vm’s to share data over low latency interlinks. This will enable applications to scale to tens of thousands of CPU cores without modification. This is now available to try out on P3dn and C5n instances.
4. AWS IoT SiteWise/AWS IoT Events offer managed services for structuring and collecting IoT device data.
AWS SiteWise is a new service that enables organizations to collect, organize, and structure data from IoT devices. Made for industrial applications this service will enable organizations to perform event and advanced analytics against IoT sensor data.
AWS IoT Events will help to monitor IoT data for malfunctioning equipment. It will trigger alerts and notifications for predictive parts failure, early warning detection, and other operationalized services.
Two other services AWS IoT Greengrass Connectors allows this data to be connected to third party applications such as ServiceNow or Splunk via API’s
Finally AWS IoT Things Graph enables a drag and drop experience for linking sensor devices to services.