News
Inferenceai& to Provide NEC with Sovereign AI Inference Platform to Accelerate AI-Native TransformationSovereignDeutsche Telekom boosts its sovereign AI push with 100 million for MistralInferenceAMD Completes Taalas Acquisition to Strengthen AI Inference PortfolioSovereignIndia’s Sarvam AI launches GPU service for open modelsClustersNew UB North Campus structure to house AI supercomputer system for research usesClustersIntroducing Together Storage Accelerator: fast, free data caching for your GPU clusterNeocloudsJensen Huang’s Charity Committed $350 Million to CoreWeaveClustersOak Ridge's AI Factory Moment: Lux, Nuclear Power, and the Next Data Center FrontierDealsNscale’s ByteDance GPU deal draws scrutiny over US chip curbsDealsNscale’s ByteDance Deal Faces Scrutiny Under US Chip CurbsInferenceai& to Provide NEC with Sovereign AI Inference Platform to Accelerate AI-Native TransformationSovereignDeutsche Telekom boosts its sovereign AI push with 100 million for MistralInferenceAMD Completes Taalas Acquisition to Strengthen AI Inference PortfolioSovereignIndia’s Sarvam AI launches GPU service for open modelsClustersNew UB North Campus structure to house AI supercomputer system for research usesClustersIntroducing Together Storage Accelerator: fast, free data caching for your GPU clusterNeocloudsJensen Huang’s Charity Committed $350 Million to CoreWeaveClustersOak Ridge's AI Factory Moment: Lux, Nuclear Power, and the Next Data Center FrontierDealsNscale’s ByteDance GPU deal draws scrutiny over US chip curbsDealsNscale’s ByteDance Deal Faces Scrutiny Under US Chip Curbs
EveryNeocloud

Intel Gaudi

Habana Gaudi 2 and 3. 1 provider on EveryNeocloud so far.

Gaudi is Intel's line of AI accelerators, designed by Habana Labs, an Israeli start-up that Intel bought in December 2019 for about $2 billion. Gaudi 2 appeared in 2022 and Gaudi 3 in 2024. Intel sold Gaudi as a cheaper alternative to Nvidia, with standard Ethernet built into each chip, but it won only a small share of the market.

The design

Each Gaudi chip has matrix engines for neural networks and many Ethernet ports on the chip itself, so servers can be linked into clusters with ordinary network switches rather than InfiniBand. Gaudi 3 carries 128 GB of HBM2e memory. Intel's software, built around PyTorch, lets many models move from GPUs with modest changes, though the range of supported tools is narrower than Nvidia's.

Who runs it

In 2025 IBM Cloud became the first cloud provider to offer Gaudi 3, in Frankfurt, Washington and Dallas, and Intel ran its own Tiber developer cloud. Naver in Korea worked with Intel on Gaudi software. Intel said in 2025 that it would not sell its planned Falcon Shores successor broadly and would focus on later rack-scale designs, leaving Gaudi 3 as its main AI chip for now.

Famous intel Gaudi providers

Further reading

Photo credits on this page

The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

The weekly digest

Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.