The go deploy team visit the AWS Summit at the Excel in London!
AWS Summit hosted at the ExCel in London is one of biggest cloud events of the year. With plenty on offer, from everything AI to advice for businesses leading in the digital space, the day was full of hands-on workshops and seminars covering all the newest and innovative Cloud technologies. It even came with an opportunity to test your time and complete a pit stop in a real F1 car!
A team from go deploy were lucky enough to attend the day, find out what the team had to say about their experience:
Jeremy Curtis – Lead Cloud Specialist at go deploy
AWS launched Lambda almost 10 years ago in 2014 and while the Serverless concept has been around prior to that it has grown considerably since then. Lambda was and still is used to connect AWS services but now there are over 250 services that can directly interconnect without using Lambda, re-emphasizing the message that Lamba should be used to transform and not just transport data. The direct integrations result in less code, lower cost, more scalable solutions and avoid the cold start latencies of Lambdas. Another message was that Serverless used to mean code without managing servers, now it means build whole applications without managing servers. There were many service/consulting partners (e.g. PwC) and one, serverlessgurus.com specifically focussing on providing AWS Serverless consulting services.
In the AWS Community theatre there were several talks on Serverless. One by Yan Cui on testing Serverless Applications using Local Stack , Lambdalith, Serverless Framework and Remocal strategies. And another by Damien Jones on Serverless Data Pipelines again emphasising that in many cases you can connect all you data processing services directly together without the using Lambda.
Jagjeet Dhuna – Cloud Specialist at go deploy
Attending the AWS summit in London (2024) has put into perspective the significant shift in technology towards AI and Machine Learning. This year the summit was a showcase of how these technologies are becoming more and more relevant just as the internet was in 1989. The summits agenda was heavily focused on generative AI, with over 80 sessions coving a broad range of AI technologies.
A personal highlight for me was AWS’s introduction to EC2 second-generation Ultra cluster. This demonstrates the dedication, investment and innovation AWS is making towards offering services which allow its customers to develop AI technologies. EC2 second-generation Ultra cluster allows developers to provision up to 20,000 H100 GPUs in a single cluster. Furthermore, its improved performance allowing 8x larger scaling and15% lower latency with the cluster.
Elastic Kubernetes Service for Machine learning was also a point of interest. A speaker from Adobe explained how they use EKS for machine learning to better train and scale their Firefly AI (AI powered image generator). Additionally, its reliance due to build in features of K8s and its ability to utilize CPU and GPUs allowing cost affective AI training development and usage.
Usama Shafique – AWS Specialist at go deploy
I had the privilege of attending the AWS Summit 2024 in London, where significant announcements were made, Amazon Bedrock is coming to the London region with the promise of unlocking great benefits for UK businesses. While AI took centre stage during the summit, it’s undeniable that the impact of good data and the importance of security are paramount. Vasquez, (Vice President of Professional Services & Gen AI Innovation Centre Amazon Web Services) underscored this crucial point, emphasizing that the effectiveness of AI tools relies heavily on robust data infrastructure and stringent security measures. Good data leads to better AI, making it smarter and more reliable in its decisions and predictions. High-quality data ensures AI performs well, providing accurate insights and solutions.
Arushi Garg, Product Manager, Amazon S3, revealed: AWS One of the significant challenges in the industry regarding big data is the cost of storage, which can skyrocket if the right services aren’t utilized. Arushi Garg, Product Manager at Amazon S3, AWS, emphasised the importance of choosing the appropriate storage class to mitigate expenses. She recommended using S3 Intelligent Tiering for data with unpredictable access patterns, regardless of object size or retention period, making it suitable for various workloads like data lakes, analytics, and user-generated content. A case study presented by Arushi Garg highlighted how illumina, a company, achieved a remarkable 60% reduction in storage costs. For every terabyte of data, illumina saved 60% on storage expenses within three months of adopting S3 Intelligent-Tiering and lifecycle policies.
During the session on “Improving Security Through Modern Application Development,” Owen Hawkins from AWS highlighted the significance of the recently launched VPC Lattice in 2023. VPC Lattice is an application networking service designed to seamlessly connect, monitor, and secure communications between services, enabling developers to focus on core business features. By simplifying API communication across accounts and addressing network administration issues, such as different IP versions or conflicting addresses, VPC Lattice provides a centralized interface for efficient service sharing across VPCs. Engineers across teams found that with VPC Lattice, they could effortlessly integrate their services with the main analytics engine, streamlining processes and meeting business needs effectively.
See you at AWS Summit 2025!