News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise Expansion
EveryNeocloud

AMD Instinct

MI250, MI300X, MI325X and MI355X. 32 providers on EveryNeocloud so far.

AMD Instinct is AMD's line of data centre accelerators and Nvidia's main GPU rival. The MI250X powers Frontier and LUMI; the MI300A, which joins CPU and GPU in one package, powers El Capitan; and the MI300X of December 2023, with 192 GB of memory, opened the line to AI clouds. The MI325X and MI355X followed in 2024 and 2025.

Memory as the selling point

AMD has led on memory size: the MI300X has 192 GB of HBM3, the MI325X 256 GB and the MI355X 288 GB. More memory lets one GPU hold a larger model, which helps inference. AMD's software, ROCm, has been the weak point compared with Nvidia's CUDA, and much of AMD's work since 2023 has gone into closing that gap with the main AI frameworks.

Who runs it

In science, AMD chips sit in three of the largest machines: El Capitan, Frontier and LUMI, as well as Setonix at Pawsey. In clouds, Microsoft Azure, Oracle and Vultr rent MI300X and newer chips, and Meta uses them for inference. In October 2025 OpenAI agreed to deploy 6 GW of AMD GPUs, starting with the MI450 in 2026, with an option to take a stake in AMD.

Famous AMD Instinct providers

Still to be written: Vultr.

Further reading

馃嚭馃嚫 United StatesFrontier The world's first exascale supercomputer, hosted at Oak Ridge National Laboratory in Tennessee. It uses AMD processors aSupercomputer 路 AMD Instinct 路 2021 Perlmutter 馃嚭馃嚫 United StatesPerlmutter A DOE supercomputer at NERSC for fusion and climate research.Supercomputer 路 Nvidia Ampere 路 2021 馃嚭馃嚫 United StatesSummit A supercomputer built by IBM for the Oak Ridge National Laboratory.Supercomputer 路 Other Nvidia GPUs 路 2018 馃嚜馃嚫 Spain and 1 moreMareNostrum Spain's main supercomputer at the Barcelona Supercomputing CenterSupercomputer 路 Nvidia Hopper 路 2005 DigitalOcean 馃嚭馃嚫 United States and 2 moreDigitalOcean Alternative cloud for developers, with GPU Droplets for AI workDeveloper cloud 路 Nvidia Hopper 路 2011 Deucalion 馃嚨馃嚬 PortugalDeucalion Portugal's fastest supercomputer, hosted at MACC in Guimar茫esSupercomputer 路 Nvidia Ampere 路 2023 馃嚭馃嚫 United StatesTitan A hybrid supercomputer built by Cray for Oak Ridge National Laboratory.Supercomputer 路 Other Nvidia GPUs 路 2012 G42 馃嚘馃嚜 United Arab Emirates and 4 moreG42 Abu Dhabi AI holding group running cloud, data centres and national compute projectsNeocloud 路 Nvidia Blackwell 路 2018 Nscale 馃嚞馃嚙 United KingdomNscale UK-based GPU cloud provider and data centre operator founded in 2024.Neocloud 路 Other Nvidia GPUs 路 2024 El Capitan 馃嚭馃嚫 United StatesEl Capitan Exascale supercomputer for US nuclear stockpile stewardship built by HPE and AMD.Supercomputer 路 AMD Instinct 路 2024 Humain 馃嚫馃嚘 Saudi ArabiaHumain Saudi AI company building data centres, cloud services and Arabic modelsNeocloud 路 Other Nvidia GPUs 路 2025 馃嚡馃嚨 JapanTsubame A long-running Japanese supercomputer series built for AI and scientific computing.Supercomputer 路 Nvidia Hopper 路 2006 Taiwania 馃嚬馃嚰 TaiwanTaiwania A series of supercomputers in Taiwan for scientific research and public access.Supercomputer 路 Other Nvidia GPUs 路 2018 Northern Data 馃嚛馃嚜 Germany and 4 moreNorthern Data German GPU cloud and data centre group for AI and high-performance computingNeocloud 路 Nvidia Hopper 路 2019 馃嚝馃嚠 Finland and 3 moreLUMI EuroHPC petascale machine in Kajaani, Finland, built by HPE on AMD hardwareSupercomputer 路 AMD Instinct 路 2023 The Stargate Project 馃嚭馃嚫 United States and 4 moreThe Stargate Project US AI infrastructure venture building data centres for OpenAISupercomputer 路 AMD Instinct 路 2025 馃嚭馃嚫 United StatesBlue Waters A US petascale research supercomputer at the University of Illinois Urbana-ChampaignSupercomputer 路 AMD Instinct 路 2012 馃嚚馃嚦 ChinaTianhe-1 Chinese petascale supercomputer at the National Supercomputing Center in TianjinSupercomputer 路 Other Nvidia GPUs 路 2009 Alice Recoque 馃嚝馃嚪 FranceAlice Recoque France's first exascale supercomputer for research and AISupercomputer 路 AMD Instinct 路 2026 Lux 馃嚭馃嚫 United StatesLux AMD-powered AI supercomputer at Oak Ridge National LabSupercomputer 路 AMD Instinct 路 2026 Discovery 馃嚭馃嚫 United StatesDiscovery AI supercomputer for the U.S. Department of EnergySupercomputer 路 AMD Instinct 路 2028 Hot Aisle WorldHot Aisle Developer-first cloud for sovereign inference on isolated AMD GPUsNeocloud 路 AMD Instinct 路 2023 AI Research Resource 馃嚞馃嚙 United KingdomAI Research Resource UK national supercomputing programme for AI research and industrySovereign cloud 路 Nvidia Hopper 路 2023 Neysa 馃嚠馃嚦 IndiaNeysa GPU cloud and AI platform for enterprises in IndiaNeocloud 路 Other Nvidia GPUs TensorWave WorldTensorWave AMD-exclusive AI cloud for training and running large modelsNeocloud 路 AMD Instinct IndiaAI Compute Portal 馃嚠馃嚦 IndiaIndiaAI Compute Portal National cloud platform for affordable AI compute in IndiaSovereign cloud 路 Nvidia Hopper Core42 馃嚘馃嚜 United Arab EmiratesCore42 Sovereign AI cloud and infrastructure provider based in the UAE.Sovereign cloud 路 Other Nvidia GPUs Crusoe 馃嚭馃嚫 United States and 3 moreCrusoe AI cloud and modular data centre operator that also builds its own power supplyNeocloud 路 Nvidia Blackwell Featherless AI 馃嚭馃嚫 United StatesFeatherless AI Serverless inference platform for open-source AI modelsInference platform 路 AMD Instinct 路 2023 Spheron Network WorldSpheron Network Marketplace for renting enterprise NVIDIA and AMD GPUsMarketplace 路 Nvidia Blackwell Hostkey 馃嚦馃嚤 Netherlands and 3 moreHostkey Dedicated GPU and bare-metal server provider in Europe and the USNeocloud 路 Other Nvidia GPUs 路 2007 SaladCloud 馃嚭馃嚫 United StatesSaladCloud Distributed GPU cloud using idle consumer hardwareNeocloud 路 Other Nvidia GPUs 路 2018

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.