AI อะไรเนี่ย

Industry

AWS and NVIDIA Deepen Strategic Collaboration to Accelerate AI

AWS and NVIDIA Deepen Strategic Collaboration to Accelerate AI

At NVIDIA GTC 2026, a significant announcement rocked the AI world: AWS and NVIDIA are deepening their strategic collaboration to supercharge AI from pilot to production. This expanded partnership is set to deploy over a million NVIDIA GPUs and introduce new Amazon EC2 instances, promising to accelerate AI development and push production workloads to new frontiers.

What Happened: A Massive Scale-Up for AI Infrastructure

The core of this collaboration lies in a monumental commitment to scale and innovate AI infrastructure. AWS will deploy over 1 million NVIDIA GPUs, encompassing the advanced Blackwell and Rubin architectures, across its global cloud regions, starting in 2026. This massive rollout is designed to meet the skyrocketing demand for AI compute, offering developers and enterprises unparalleled access to cutting-edge hardware.

A key highlight is Amazon EC2's pioneering support for NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs. AWS is set to be the first major cloud provider to integrate these powerful GPUs, with new EC2 instances built on the robust AWS Nitro System. These instances are perfectly suited for a diverse range of compute-intensive tasks, including data analytics, conversational AI, content generation, and various graphics workloads, ensuring high performance and enhanced security.

Innovations in Interconnect and Data Processing

Beyond raw compute power, the collaboration focuses on optimizing the entire AI development and deployment lifecycle. AWS announced support for the NVIDIA Inference Xfer Library (NIXL) with AWS Elastic Fabric Adapter (EFA). This integration is crucial for accelerating disaggregated Large Language Model (LLM) inference on Amazon EC2, working seamlessly across NVIDIA GPUs and AWS Trainiums. It also integrates natively with popular open-source frameworks like NVIDIA Dynamo, vLLM, and SGLang, promising faster, more efficient inference.

For data professionals, the news is equally exciting. Apache Spark performance is set to become 3x faster using Amazon EMR on Amazon Elastic Kubernetes Service (Amazon EKS) with Amazon EC2 G7e instances, which are powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. This significant boost will help data engineers and scientists accelerate their data processing pipelines, cutting down time-to-insight for AI/ML feature engineering and complex ETL transformations.

Expanding Model Capabilities on Amazon Bedrock

The partnership also extends to enhancing model development and fine-tuning. Expanded NVIDIA Nemotron model support is coming to Amazon Bedrock, AWS's fully managed service that provides access to foundation models. This will allow developers to fine-tune NVIDIA Nemotron models directly on Amazon Bedrock using Reinforcement Fine-Tuning (RFT), making it easier to align model behavior with specific domain requirements without infrastructure overhead. Additionally, NVIDIA Nemotron 3 Super, a hybrid MoE model designed for multi-agent workloads and extended reasoning, will also be available on Amazon Bedrock, enabling sophisticated AI agents for various industries.

Why This Matters for Builders

This deepened collaboration is a game-changer for anyone building or deploying AI solutions. It translates directly into:

  • Unprecedented Access to AI Hardware: Developers and enterprises will have immediate access to the latest and most powerful NVIDIA GPUs, significantly lowering the barrier to entry for advanced AI projects.
  • Faster Development and Deployment: Optimized interconnects, accelerated data processing, and integrated model support mean quicker experimentation, faster training, and more efficient inference for large-scale AI models.
  • Scalability and Production Readiness: The focus is firmly on moving AI from experimental pilots to robust production environments, ensuring reliability, performance at scale, and adherence to security and compliance standards.
  • Innovation in Agentic AI: With new models and enhanced infrastructure, the collaboration empowers the development of agentic AI systems capable of complex reasoning and autonomous actions across workflows.

This strategic move by AWS and NVIDIA underscores their commitment to pushing the boundaries of what's possible in AI, providing the foundational tools necessary for the next wave of innovation. For more details on this extensive collaboration, you can read the official announcement.

Read more:

Learn more about how AWS and NVIDIA Deepen AI Collaboration to accelerate AI innovation.