Intel® Gaudi® 3 AI Accelerators
Bringing choice to gen AI with performance, scalability and efficiency. Meet your new high-performance option built to handle your AI workloads, your way.
Intel® Gaudi® 3 AI Accelerators
A Better Choice for Enterprise AI
Built on the high-efficiency Intel® Gaudi® platform with proven MLPerf benchmark performance, Intel Gaudi 3 AI accelerators are built to handle demanding training and inference.
Support AI applications like Large Language Models, Multi-Modal Models and Enterprise RAG in your data center or in the cloud—from node to mega cluster, all running on the Ethernet infrastructure you likely already own. Whether you need a single accelerator or thousands, Intel Gaudi 3 can play a pivotal role in your AI success.
Open
Avoid risky investments in locked, proprietary technologies such as NVLink, NVSwitch, and InfiniBand.
Built for Ethernet
Use the networking infrastructure you already own and support future needs with standard Ethernet hardware.
More I/O
Take advantage of 33 percent more I/O connectivity per accelerator compared to H100,1 so you can enable massive scale-up and scale-out with optimized cost efficiency.
Scale Effortlessly
Intel® Gaudi 3 AI Accelerators are designed to deliver simple, cost-effective AI scalability for even the largest and most complex deployments.
Intel Gaudi 3 AI Accelerator
32-Node Cluster Reference Design
Build out AI solutions with the latest Intel Gaudi 3 accelerator-based systems—built for scale and expandability with all-Ethernet-based fabrics and support for a wide range of industry AI models and frameworks.
Designed for the Real-World Demands of AI
Intel® Gaudi® 3 AI accelerators empower you to use open, community-based software and industry-standard Ethernet networking to scale systems more flexibly.
Save Time and Power with Intel Gaudi 3 Accelerators
2x
AI compute (FP8) vs. Intel Gaudi 2 AI Accelerators
4x
AI compute (BF16) vs. Intel Gaudi 2 AI Accelerators
2x
Network bandwidth vs. Intel Gaudi 2 AI Accelerators
2x
AI compute (FP8) vs. Intel Gaudi 2 AI Accelerators
4x
AI compute (BF16) vs. Intel Gaudi 2 AI Accelerators
2x
Network bandwidth vs. Intel Gaudi 2 AI Accelerators
Develop with Ease
Designed for rapid ROI. Getting started with Intel Gaudi 3 AI accelerators is simple—regardless of whether you’re starting from scratch, fine-tuning off-the-shelf models, or migrating from a GPU-based approach.
Optimized for Developers
Take advantage of software tools and developer resources to get up to speed effortlessly.
Support for New and Existing Models
Customize reference models, start fresh, or migrate existing models using open source tools, including resources from Hugging Face.
Integrated with PyTorch
Keep working with the library your team already knows.
Easy Migration of GPU-Based Models
Quickly port your existing solutions using our purpose-built software tools.
Simplify Development End to End
Go from proof of concept to production in less time. From migration to deployment, Intel Gaudi 3 AI Accelerators come supported by a powerful portfolio of software tools, resources, and training. Get to know what’s available to help simplify your AI efforts.
Deploy Intel Gaudi 3 AI Accelerators via OEM Designs
We’re working with our OEM partners to bring Intel Gaudi 3 AI accelerators to on-prem deployments. For more information on purchasing, please reach out to your OEM partner or Intel representative.
Experience Intel Gaudi AI Accelerators in the Cloud
Invest with Confidence
Join our mission for AI innovation. When you buy Intel Gaudi 3 AI accelerators, you join a continuing advancement road map that will help steward AI into the future and unlock full-scale deployments across industries.
Explore Intel Gaudi 3 Products and Solutions
Get the Latest on AI Trends and Technologies
Subscribe to stay connected with Intel
Product and Performance Information
NVIDIA H100 GPU (900 GB/s closed NVLink connectivity) vs. Intel® Gaudi® 3 accelerator (1200 GB/s open standard RoCE).