- AWS›
- Global Infrastructure›
- AI Factories
AWS AI Factories
AWS-built AI infrastructure. Ready to use in your data center. Deployed fast.
Fully managed AI infrastructure in your data centers
You provide a data center and power, we'll deploy a dedicated, secure, and fully managed AI infrastructure for you. Leveraging nearly two decades of AWS cloud expertise, AI Factories help you eliminate years of build effort and maintenance.
Built for Secure & Sovereign AI
Meet data sovereignty and security requirements with dedicated deployments built exclusively for you. You can train and run large models on proprietary data while meeting strict regulatory requirements for where data is processed and stored.
Innovate and Scale AI Workloads
Innovate with services like Amazon EC2, Bedrock, and Amazon SageMaker while AWS deploys and operates the underlying infrastructure, including the latest AI chips including Trainium accelerators and NVIDIA GPUs, networking, and storage.
Features
Latest AI Chips at Massive Scale
Run demanding AI workloads on EC2 instances accelerated by the latest Trainium accelerators including Trainium2 and Trainium3 as well as NVIDIA GPUs including B200, GB200, and upcoming B300, GB300, delivering the highest GPU performance for large-scale training and inference.
Petabit-Scale Network Fabric
Deploy EC2 UltraClusters interconnected using Elastic Fabric Adapter (EFA) networking in a petabit-scale non-blocking network, helping you scale to thousands of GPUs and access to several exaflops of accelerated compute.
High-performance Storage
Using Amazon FSx for Lustre and Amazon S3 Express One Zone, you can access data at hundreds of GBp/s of throughput and millions of IOPS required for large-scale AI model training and inference workloads.
Enterprise AI Services & Models
Build AI applications with your proprietary data using specialized AI services, including Amazon Bedrock and Amazon SageMaker, to access leading foundation models as well as, build, train, and deploy your own AI models.
Leverage Existing Investments
Leverage your data center space and power you've already invested in, with the option to also use your NVIDIA GPUs, meeting you wherever you are in your AI development journey.
Enhanced Security and Stability
Secure sensitive workloads with AWS Nitro System’s hardware and firmware designed to enforce restrictions so that no one, including AWS, can access your sensitive AI workloads. The Nitro System deploys firmware updates and bug fixes while staying operational and increasing stability.
Use Cases
Accelerate AI-powered applications
Deploy AI inference applications using Amazon SageMaker and Bedrock. AWS provides specialized networking, high-performance storage, and comprehensive AI services, transforming your infrastructure into AI environments.
Sovereign AI Computing
Meet data sovereignty and compliance requirements through dedicated environments enabling regulated industries and governments to harness AI's power while adhering to security needs and advancing economies.
Training Frontier AI Models
AWS AI Factories deliver specialized compute infrastructure to train advanced AI models of all different sizes. With NVIDIA GPUs and UltraCluster, process massive datasets to build models that tackle complex tasks with speed to maximize training performance at scale.
What's new with AWS AI Factories
Learn more about AWS AI Factories
Contact your AWS account team to learn more about deploying AWS AI Factories in your data center today!
