Intel® Core™ Ultra Processors
Learn More about Intel

Intel® Gaudi® 3 AI Accelerator

For leading price performance for the cloud and on-premises

The Intel® Gaudi® 3 AI accelerator delivers best-in-class performance for your most demanding GenAI and LLM training and inference workloads. Built on the high-performance, high-efficiency Gaudi platform with proven performance on MLPerf benchmark.

Intel® Gaudi® 3: A High-Performance AI Accelerator

Intel® Gaudi® 3 is your high-performance AI accelerator option for every type of enterprise AI workload.

High Performance

Delivers best-in-class performance for your most demanding GenAI and LLM workloads.

Scalability

Flexibility to efficiently scale from 1 to 100s of nodes to meet your GenAI and LLM requirements.

Focused Architecture

Giving you choice by removing barriers, making it easy to migrate models from other GPUs.

Coming Soon: Dell PowerEdge XE9680 with Intel Gaudi 3 AI Accelerator

Discover the key highlights that make the Intel Gaudi 3 AI accelerator superior to traditional GPU offerings in the market.
Learn More
With enterprise-ready performance, scalability, and efficiency that gives more choice to more customers, Intel® Gaudi® 3 accelerators help unlock insight, innovation, and income opportunities for your organization.

Supercharge Your AI Use Case

Make the amazing possible at any scale. Intel® Gaudi® 3 AI accelerators support state-of-the-art generative AI and LLMs in the data center and pair with Intel® Xeon® processors—the host CPU of choice for leading AI systems to deliver enterprise performance and reliability.
3D Generation
Sentiment Analysis
Text-to-Image Generation
Language Translation
Text Classification
Video Retrieval
Question Answering
Text Generation and Summarization
Visual Question Answering

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.
Built for training and inference

Built for training and inference

64

Tensor Processor Cores (5th gen)

8

Matrix math engines
Increased memory for LLM efficiency and cost-effectiveness

Increased memory for LLM efficiency and cost-effectiveness

128 GB

HBM capacity, 3.7 TB/s B/W

96 MB

SRAM, 12.8 TB/s SRAM B/W
Massive, flexible on-chip networking

Massive, flexible on-chip networking

Open standard vs. proprietary InfiniBand

24x 200GbE

Industry-standard RoCE Ethernet ports

PCle 5

x 16

See the videos and learn more about Intel® Gaudi® 3 AI Accelerator

AI is Reshaping our World
How AI is reshaping our world.

Ease of adoption

Intel Gaudi 3
Intel Gaudi 3

Ethernet

Intel Gaudi 3 AI Accelerator
Intel Gaudi 3 AI Accelerator

Open Standards

Intel Gaudi

Try Intel® Gaudi® AI Accelerators

Put our dedicated AI platform to the test. You can experience Intel® Gaudi® AI accelerators for yourself in the Intel® Tiber™ Developer Cloud.
Try it now
Intel® Core™ Ultra Processors
Learn More about Intel