AWS re:Invent 2024: Keynote Highlights and Technical Updates

Friday 6th December 2024

Keynote Highlights

Dr. Ruba Borno, Vice President Global Specialists – ‘AWS unveils new partner innovations at re:Invent 2024’

Dr. Ruba Borno's keynote on Wednesday highlighted significant innovations in cloud security and partner collaborations. Key announcements included:

  • New Security Specialisations: AWS launched four new security specializations, including the Security AI Category, to help customers identify partners with deep expertise in protecting AI applications and defending against cyberattacks.
  • Buy with AWS: Now generally available, this feature allows partners to integrate AWS Marketplace directly into their websites, enabling customers to discover solutions, request pricing, and make purchases using their AWS accounts. Notably, Bytes is the only EMEA launch partner and one of 17 globally. Read the full AWS news blog: Introducing Buy with AWS: an accelerated procurement experience on AWS Partner sites, powered by AWS Marketplace
  • Partner Connections: Announced in preview, this unified co-sell experience simplifies deal structures across multiple partners on a single customer project.
  • Migration Acceleration Program Enhancements: Enhancements include the removal of earnings caps on modernisation deals.

Try it now! Buy with AWS

Swami Sivasubramanian, Vice President of AI and Data – ‘AWS announces powerful new capabilities for Amazon Bedrock and biggest expansion of models to date’

Swami Sivasubramanian's keynote on December 4th introduced a range of innovations for Amazon Bedrock, a fully managed service for building and scaling generative AI applications. Highlights include:

  • Broad Selection of Models: Access to over 100 popular, emerging, and specialised models through Amazon Bedrock Marketplace.
  • New Capabilities: Enhancements to manage prompts at scale, new capabilities for Amazon Bedrock Knowledge Bases, and Amazon Bedrock Data Automation.
  • Availability: Amazon Bedrock Marketplace is available today, with other capabilities in preview. Models from Luma AI, Poolside, and Stability AI are coming soon.

Technical Updates

1. Analytics

  • Amazon SageMaker Lakehouse: Simplifies analytics and AI/ML by unifying data silos, integrating S3 data lakes and Redshift warehouses for unified analytics and AI/ML on a single data copy.
  • Amazon Q in QuickSight: Enables complex scenario analysis up to 10x faster than spreadsheets.
  • Amazon DynamoDB Zero-ETL Integration: Allows seamless analysis of operational data in Amazon SageMaker Lakehouse.
  • AWS Clean Rooms: Now supports multiple clouds and data sources, enhancing secure data collaboration.

2. Compute

  • Amazon EC2 Trn2 Instances and Trn2 UltraServers: Offer 4x faster speed, 4x more memory bandwidth, and 30% higher floating-point operations for AI/ML training and inference.
  • Amazon EC2 P5en Instances: Feature NVIDIA H200 Tensor Core GPUs and EFAv3 networking, delivering up to 3,200 Gbps network bandwidth.
  • Amazon EC2 I8g and I7ie Instances: Provide optimised storage performance for I/O-intensive workloads.

3. Containers

  • Amazon EKS Hybrid Nodes: Enable the use of on-premises infrastructure in Amazon EKS clusters.
  • Amazon EKS Auto Mode: Simplifies Kubernetes cluster management by automating compute, storage, and networking.

4. Database

  • Amazon MemoryDB Multi-Region: Now generally available, offering microsecond latencies across regions and up to 99.999% availability.

5. Generative AI / Machine Learning

  • Next Generation of Amazon SageMaker: Unifies data engineering, analytics, and generative AI in a streamlined studio.
  • Amazon Q Developer: Empowers non-ML experts to build ML models using natural language.
  • Amazon Bedrock Guardrails: Supports multimodal toxicity detection with image support.
  • Amazon Bedrock Enhancements: Include multimodal processing, graph modeling, and structured querying.
  • Amazon Bedrock Marketplace: Access over 100 foundation models in one place.
  • Amazon SageMaker HyperPod: Offers flexible training plans and task governance for efficient model development.
  • Amazon SageMaker Data and AI Governance: Provides a unified catalog and granular access controls.
  • Amazon Q Business: Adds workflow automation capabilities and 50+ action integrations.
  • Amazon Nova Foundation Models: Deliver frontier intelligence and industry-leading price performance.
  • Amazon Bedrock Multi-Agent Collaboration: Allows developers to build and manage multiple specialised agents.
  • Automated Reasoning Checks: Enhance conversational AI accuracy by reducing hallucinations.
  • Amazon Bedrock Model Distillation: Automates the creation of distilled models for specific use cases.
  • RAG Evaluation and LLM-as-a-Judge: Offer quality and responsible AI metrics at scale.
  • Amazon Q Business Extensions: Provide seamless AI assistance within work applications.
  • Amazon Bedrock APIs: Enhance RAG applications with custom connectors and reranking models.
  • PartyRock Capabilities: Include image generation, visual analysis, and multi-doc processing.
  • Amazon Q Business Visual Insights: Extract insights from visual elements within documents.

6. Migration & Transfer Services

  • AWS Database Migration Service: Automates schema conversion tasks using generative AI.
  • AWS Transfer Family Web Apps: Provide a simple interface for accessing data in Amazon S3.
  • Amazon S3 Data Integrity Protections: Introduce default data integrity protections for new objects.

AWS re:Invent 2024 has showcased a plethora of innovations, reinforcing AWS's commitment to advancing technology and supporting its global community. Stay tuned for more updates as the event continues to unfold!

Bytes have various workshops, assessments and funding options to support you on your AWS journey. As an AWS Premier Tier Services Partner - we're in the best position to help you! Email [email protected] today.


Want to keep informed? Sign up to our Newsletter

Connect