Exclusive Content:

Haiper steps out of stealth mode, secures $13.8 million seed funding for video-generative AI

Haiper Emerges from Stealth Mode with $13.8 Million Seed...

Running Your ML Notebook on Databricks: A Step-by-Step Guide

A Step-by-Step Guide to Hosting Machine Learning Notebooks in...

“Revealing Weak Infosec Practices that Open the Door for Cyber Criminals in Your Organization” • The Register

Warning: Stolen ChatGPT Credentials a Hot Commodity on the...

The process of how infrastructure is engineered by AWS engineers to support generative AI applications

Revolutionizing Networking for Generative AI: Introducing UltraCluster 2.0

Delivering low-latency, large-scale networking is key to enabling the training and running of complex generative AI models efficiently. With the demand for AI-powered applications on the rise, it’s crucial for tech companies to invest in networks that can handle the scale and speed required for these workloads.

At AWS, we understand the importance of reducing network latency and enhancing performance for our customers. That’s why we have taken a unique approach by building our own network devices and operating systems for every layer of the stack. This level of control not only allows us to improve security, reliability, and performance but also enables us to innovate at a faster pace.

One of our key innovations in this space was the introduction of the Elastic Fabric Adapter (EFA) in 2019. This custom-built network interface provides operating system bypass capabilities to Amazon EC2 instances, allowing customers to run applications with high levels of inter-node communication at scale. EFA utilizes the Scalable Reliable Datagram (SRD) protocol, designed specifically by AWS, to deliver high performance and low latency.

Most recently, we launched the UltraCluster 2.0 network, designed specifically for generative AI workloads. This upgraded network supports over 20,000 GPUs with a 25% reduction in latency compared to its predecessor. Built in just seven months, UltraCluster 2.0 delivers tens of petabits per second of throughput with a round-trip time of less than 10 microseconds. This new network has resulted in at least a 15% reduction in model training time, showcasing its efficiency and effectiveness for large-scale AI projects.

By investing in custom-built network devices and software, we have been able to develop innovative solutions that meet the demands of AI workloads. Our commitment to delivering low-latency, large-scale networking solutions continues to drive advancements in the field of artificial intelligence and propel our customers towards groundbreaking discoveries and innovations.

Latest

Reinforcement Fine-Tuning for Amazon Nova: Educating AI via Feedback

Unlocking Domain-Specific Capabilities: A Guide to Reinforcement Fine-Tuning for...

Calculating Your AI Footprint: How Much Water Does ChatGPT Consume?

Understanding the Hidden Water Footprint of AI: Balancing Innovation...

China’s AI² Robotics Secures $145M in Funding for Model Development and Humanoid Robot Enhancements

AI² Robotics Secures $145 Million in Series B Funding...

Don't miss

Haiper steps out of stealth mode, secures $13.8 million seed funding for video-generative AI

Haiper Emerges from Stealth Mode with $13.8 Million Seed...

Running Your ML Notebook on Databricks: A Step-by-Step Guide

A Step-by-Step Guide to Hosting Machine Learning Notebooks in...

VOXI UK Launches First AI Chatbot to Support Customers

VOXI Launches AI Chatbot to Revolutionize Customer Services in...

Investing in digital infrastructure key to realizing generative AI’s potential for driving economic growth | articles

Challenges Hindering the Widescale Deployment of Generative AI: Legal,...

How AI is Transforming Cybersecurity

Navigating the Dual Challenge of AI: Evolving Threats and Strategic Cyber Defense This heading encapsulates the complex interplay between the challenges posed by AI's rapid...

Transforming Observability with Generative AI and OpenTelemetry

Generative AI Adoption Surges to 98% as OpenTelemetry Redefines Production Environments by David Hope, February 18, 2026 Explore how generative AI and OpenTelemetry are revolutionizing...

What is the Impact of Generative AI on Science?

The Dawn of AI Collaboration in Scientific Research: A New Chapter in Authorship? The New Era of AI in Scientific Research: A Double-Edged Sword In February...