Amazon Web Services made a lot of announcements at the AWS re:Invent event in Las Vegas. We have summarised everything that Amazon announced.
Amazon Web Services is one of the technology giant’s main sources of revenue and provides the infrastructure for many internet services. AWS is known as the world’s most comprehensive and popular cloud service.
To maintain this advantage, Amazon holds an event called AWS re:Invent in Las Vegas to announce its innovations. In the first two days of the event, many new features were announced. We have summarised those features for you.
Announcements from the first day
Amazon Q
Amazon Q, the biggest announcement of the day, is an artificial intelligence chatbot designed to serve AWS customers. AWS CEO Adam Selipsky explained that this bot can easily chat, produce content and take action. This customisable bot will also be accessible from different platforms. In particular, operational processes can be simplified with Amazon Q.
Guardrails for Amazon Bedrock
With Guardrails for Amazon Bedrock, organisations can define the language a model will use and apply restrictions. For example, this model will prevent artificial intelligence chat bots from answering questions on topics unrelated to the company’s work.
New chips: AWS Trainium2 and Graviton4
Amazon also introduced two new chips at the event. The first of these chips is Trainium2, which will replace the first generation Trainium, which has twice the electricity consumption and four times better performance than the previous Trainium chips. The other chip is Graviton4, which will be used in inference operations. This chip has 30 per cent better performance, 50 per cent more cores and 75 per cent higher memory bandwidth. Both chips will be used in artificial intelligence studies.
Amazon S3 Express Single Zone
Amazon has announced a major enhancement to its S3 object storage service. With S3 Express One Zone, it will be possible to achieve 10 times the performance of the standard S3 object storage service. This feature will be a huge advantage for artificial intelligence training and computations that require high capacity.
Virtual desktop environment
Amazon’s new $195 tool prepares virtual desktop environments. This virtual desktop on Fire TV Cube hardware enables companies to work without investing in large technology infrastructures.
Neptune Analytics
Neptune Analytics, another tool introduced by AWS, enables the combination of graph and vector databases. With this tool, it is possible to make sense of the data in S3 object stores.
Clean Rooms ML
With Clean Rooms ML, Amazon’s customers will be able to use company-specific trained artificial intelligence models that emphasise privacy and are similar to each other. Companies will be able to collaborate on this platform. With Clean Rooms ML, there will be no need to share data with external partners.
SageMaker HyperPod
With SageMaker HyperPod, it will be possible to train and tune large artificial intelligence language models on the cloud computing side of AWS. Thanks to this tool, it will be possible to train models in an accelerated manner over distributed clusters and test models and data simultaneously on different clusters.
AWS Titan Visual Builder
