Nebius to Offer NVIDIA Blackwell Ultra-Powered Instances
Nebius (NASDAQ:NBIS) announces its role as an early adopter cloud provider of the NVIDIA Blackwell Ultra platform, offering NVIDIA GB300 NVL72-powered instances by the end of 2025. The instances will be accelerated by 72 NVIDIA Blackwell Ultra GPUs.
The company plans to make NVIDIA Blackwell GPU capacity generally available at its U.S. data centers starting in Q2 2025. Nebius's New Jersey data center will be dedicated exclusively to NVIDIA Blackwell-architecture GPUs, while its Kansas City facility will deploy NVIDIA HGX B200.
Additionally, Nebius has been named as an ecosystem partner for the NVIDIA Dynamo Inference Engine, an open-source framework for deploying generative AI that increases DeepSeek-R1 throughput by 30x on NVIDIA Blackwell. Pre-orders for NVIDIA GB200 NVL72 and NVIDIA HGX B200 instances are currently available on Nebius's website.
Nebius (NASDAQ:NBIS) annuncia il suo ruolo come fornitore di cloud early adopter della Piattaforma NVIDIA Blackwell Ultra, offrendo istanze alimentate da NVIDIA GB300 NVL72 entro la fine del 2025. Le istanze saranno accelerate da 72 GPU NVIDIA Blackwell Ultra.
L'azienda prevede di rendere disponibile la capacità GPU NVIDIA Blackwell nei suoi data center negli Stati Uniti a partire dal secondo trimestre del 2025. Il data center del New Jersey sarà dedicato esclusivamente alle GPU con architettura NVIDIA Blackwell, mentre la struttura di Kansas City implementerà NVIDIA HGX B200.
Inoltre, Nebius è stata nominata partner ecosistemico per il NVIDIA Dynamo Inference Engine, un framework open-source per l'implementazione di AI generativa che aumenta il throughput di DeepSeek-R1 di 30 volte su NVIDIA Blackwell. Gli ordini anticipati per le istanze NVIDIA GB200 NVL72 e NVIDIA HGX B200 sono attualmente disponibili sul sito web di Nebius.
Nebius (NASDAQ:NBIS) anuncia su papel como proveedor de nube pionero de la plataforma NVIDIA Blackwell Ultra, ofreciendo instancias impulsadas por NVIDIA GB300 NVL72 para finales de 2025. Las instancias estarán aceleradas por 72 GPUs NVIDIA Blackwell Ultra.
La empresa planea hacer disponible la capacidad GPU NVIDIA Blackwell en sus centros de datos en EE. UU. a partir del segundo trimestre de 2025. El centro de datos de Nueva Jersey estará dedicado exclusivamente a GPUs de arquitectura NVIDIA Blackwell, mientras que su instalación en Kansas City desplegará NVIDIA HGX B200.
Además, Nebius ha sido nombrada socio del ecosistema para el NVIDIA Dynamo Inference Engine, un marco de código abierto para implementar IA generativa que aumenta el rendimiento de DeepSeek-R1 en 30 veces en NVIDIA Blackwell. Los pedidos anticipados para las instancias NVIDIA GB200 NVL72 y NVIDIA HGX B200 están actualmente disponibles en el sitio web de Nebius.
네비우스(NASDAQ:NBIS)는 NVIDIA 블랙웰 울트라 플랫폼의 초기 채택 클라우드 제공업체로서의 역할을 발표하며, 2025년 말까지 NVIDIA GB300 NVL72로 구동되는 인스턴스를 제공할 예정입니다. 이 인스턴스는 72개의 NVIDIA 블랙웰 울트라 GPU에 의해 가속화됩니다.
회사는 2025년 2분기부터 미국 데이터 센터에서 NVIDIA 블랙웰 GPU 용량을 일반적으로 제공할 계획입니다. 뉴저지의 데이터 센터는 NVIDIA 블랙웰 아키텍처 GPU 전용으로 운영되며, 캔자스 시티 시설에서는 NVIDIA HGX B200을 배포할 것입니다.
또한, 네비우스는 NVIDIA 다이나모 추론 엔진의 생태계 파트너로 선정되어, NVIDIA 블랙웰에서 DeepSeek-R1의 처리량을 30배 증가시키는 생성 AI 배포를 위한 오픈 소스 프레임워크를 제공합니다. NVIDIA GB200 NVL72 및 NVIDIA HGX B200 인스턴스의 사전 주문은 현재 네비우스 웹사이트에서 가능합니다.
Nebius (NASDAQ:NBIS) annonce son rôle en tant que fournisseur de cloud précurseur de la plateforme NVIDIA Blackwell Ultra, offrant des instances propulsées par NVIDIA GB300 NVL72 d'ici la fin de 2025. Les instances seront accélérées par 72 GPU NVIDIA Blackwell Ultra.
L'entreprise prévoit de rendre la capacité GPU NVIDIA Blackwell généralement disponible dans ses centres de données aux États-Unis à partir du deuxième trimestre 2025. Le centre de données du New Jersey sera exclusivement dédié aux GPU architecturés NVIDIA Blackwell, tandis que son installation à Kansas City déploiera NVIDIA HGX B200.
De plus, Nebius a été désignée comme partenaire de l'écosystème pour le NVIDIA Dynamo Inference Engine, un cadre open-source pour le déploiement d'IA générative qui augmente le débit de DeepSeek-R1 de 30 fois sur NVIDIA Blackwell. Les précommandes pour les instances NVIDIA GB200 NVL72 et NVIDIA HGX B200 sont actuellement disponibles sur le site web de Nebius.
Nebius (NASDAQ:NBIS) gibt seine Rolle als Early-Adopter-Cloud-Anbieter der NVIDIA Blackwell Ultra-Plattform bekannt und bietet bis Ende 2025 instanzen an, die von NVIDIA GB300 NVL72 betrieben werden. Die Instanzen werden durch 72 NVIDIA Blackwell Ultra GPUs beschleunigt.
Das Unternehmen plant, die NVIDIA Blackwell GPU-Kapazität ab dem 2. Quartal 2025 in seinen Rechenzentren in den USA allgemein verfügbar zu machen. Das Rechenzentrum in New Jersey wird ausschließlich NVIDIA Blackwell-Architektur-GPUs gewidmet sein, während die Einrichtung in Kansas City NVIDIA HGX B200 einsetzen wird.
Zusätzlich wurde Nebius als Ökosystempartner für die NVIDIA Dynamo Inference Engine benannt, ein Open-Source-Framework zur Bereitstellung generativer KI, das den Durchsatz von DeepSeek-R1 auf NVIDIA Blackwell um das 30-fache erhöht. Vorbestellungen für die NVIDIA GB200 NVL72 und NVIDIA HGX B200 Instanzen sind derzeit auf der Website von Nebius verfügbar.
- Early access to cutting-edge NVIDIA Blackwell Ultra technology positions company as leading AI infrastructure provider
- Dedicated data center in New Jersey exclusively for NVIDIA Blackwell-architecture GPUs shows significant infrastructure investment
- 30x throughput improvement with NVIDIA Dynamo implementation demonstrates superior performance capabilities
- Strategic partnership with NVIDIA strengthens market position in AI cloud services
- Significant capital investment required for new GPU infrastructure may impact short-term financials
- Extended timeline to full deployment (end of 2025) may allow competitors to catch up
- Heavy reliance on single hardware provider (NVIDIA) poses potential supply chain risks
Insights
Nebius's early adoption of NVIDIA's Blackwell Ultra platform represents a significant strategic positioning in the rapidly evolving AI cloud infrastructure landscape. The GB300 NVL72-powered instances with 72 Blackwell Ultra GPUs represent the absolute cutting edge of AI compute capability—hardware specifically architected for the most demanding AI workloads including reasoning models and AI agents.
The technical differentiation here is substantial. By dedicating an entire New Jersey data center to Blackwell architecture GPUs and deploying HGX B200 in Kansas City with availability beginning Q2 2025, Nebius is building specialized AI infrastructure rather than simply adding AI capabilities to general-purpose cloud environments. This purpose-built approach typically yields superior performance for AI workloads.
Particularly notable is Nebius's partnership on NVIDIA Dynamo, which provides 30x throughput improvement for inference on models like DeepSeek-R1. This inference optimization capability is critical as it directly impacts the economics of running large AI models at scale—potentially allowing Nebius customers to achieve significantly better price-performance than competitors using less optimized infrastructure.
The timeline—with Blackwell availability starting next quarter and the more advanced NVL72 configurations by year-end—gives Nebius a key technical advantage during a period when demand for advanced AI compute continues to outstrip supply. For AI companies developing next-generation applications requiring massive compute resources, early access to this infrastructure could be the difference between bringing innovative products to market quickly or facing development delays.
Nebius's strategic move to become an early adopter of NVIDIA's Blackwell Ultra platform strengthens its competitive positioning in the high-margin, high-growth AI infrastructure space. By securing early access to NVIDIA's most advanced GPU architecture, Nebius addresses a critical market bottleneck—access to specialized AI compute—that continues to constrain AI development across industries.
The dedicated GPU data centers in New Jersey and Kansas City represent significant capital investments that should be evaluated against potential returns from high-value AI workloads. While exact investment figures aren't disclosed, these specialized facilities typically command premium pricing compared to general-purpose cloud infrastructure, potentially improving Nebius's revenue per data center square foot and overall margin profile.
The Q2 2025 availability timeline for initial Blackwell capacity creates a near-term catalyst for Nebius, potentially accelerating customer acquisition among AI startups and enterprises seeking scarce compute resources. The company's offering of pre-orders for these instances could generate advance commitments and improve capacity utilization forecasting.
Particularly valuable is Nebius's ecosystem partnership for NVIDIA Dynamo, which enhances inference efficiency—the most frequent and potentially most profitable GPU workload type. The 30x throughput improvement mentioned for DeepSeek-R1 models could translate to meaningful operating cost advantages for Nebius customers, strengthening customer retention and potentially expanding margins.
This infrastructure investment positions Nebius to capture growing demand from both AI-native companies and enterprises seeking to build and deploy advanced AI applications without the massive upfront capital expenditure of building their own specialized infrastructure.
- Nebius will be an early adopter cloud provider of the NVIDIA Blackwell Ultra platform, offering NVIDIA GB300 NVL72-powered instances
-
Announces general availability of NVIDIA Blackwell GPU capacity at
U.S. data centers beginning in Q2 - Also named as ecosystem partner for new NVIDIA Dynamo Inference Engine
In support of the launch of NVIDIA Blackwell Ultra today at GTC 2025, Nebius will give its customers access to NVIDIA GB300 NVL72-powered instances – accelerated by 72 NVIDIA Blackwell Ultra GPUs – by the end of 2025.
Arkady Volozh, founder and CEO of Nebius, said:
“Access to world-class AI infrastructure is the key to realizing the full potential of AI. With dedicated NVIDIA Blackwell capacity available from next quarter, Nebius is giving AI innovators and enterprises everywhere access to the world’s most powerful AI compute through our AI-native multi-tenant cloud. This is the future of AI, and we’re building it today.”
“The NVIDIA Blackwell platform was built to bring the power of AI to enterprises and innovators worldwide,” said Dave Salvator, director of accelerated computing products at NVIDIA. “As one of the first GPU cloud providers to offer NVIDIA Blackwell and Blackwell Ultra-powered instances, Nebius will enable its customers to accelerate next-generation reasoning models, AI agents and physical AI applications.”
Today Nebius was also confirmed as an ecosystem partner for NVIDIA Dynamo, an open-source inference serving framework for deploying generative AI in large-scale distributed environments. NVIDIA Dynamo provides the most efficient solution for scaling test-time compute. Dynamo on NVIDIA Blackwell boosts throughput on DeepSeek-R1 by 30x.
Nebius today confirmed that its recently announced data center in
Pre-orders for NVIDIA GB200 NVL72 and NVIDIA HGX B200 instances can be placed now on Nebius’s website.
Nebius is currently exhibiting at GTC 2025, demonstrating the capabilities of its AI Cloud and AI Studio for AI builders, innovators and enterprises at Booth 809.
About Nebius
Nebius is a technology company building full-stack cloud infrastructure to serve the explosive growth of the global AI industry, including large-scale GPU clusters, cloud platforms, and tools and services for developers. Headquartered in
Nebius’ core business is an AI-native cloud platform that delivers a true hyperscale cloud experience tailored for AI innovators. With proprietary cloud software architecture and hardware designed in-house (including servers, racks and data center design), Nebius gives AI builders the compute, storage, managed services and tools they need to build, tune and run models and apps in one place.
A Reference Platform NVIDIA Cloud Partner, Nebius builds its AI Cloud on NVIDIA Reference Architecture, verified through adoption, strict adherence, and validation of NVIDIA Cloud Partner Reference Architecture across the entire technology stack of hardware and software components.
Nebius holds Reference Platform NVIDIA Cloud Partner status, one of only a handful of AI cloud providers to meet these requirements globally, underscoring Nebius’ expertise in designing and deploying a full stack of hardware and software infrastructure that meets NVIDIA’s standards for AI and ML workloads.
To learn more please visit www.nebius.com
Disclaimer
Forward Looking Statements
This press release and the materials referenced herein contain forward-looking statements that involve risks and uncertainties. All statements contained or implied other than statements of historical facts, including, without limitation, statements regarding our business plans, market opportunities, capacity build-out plans, capital expenditure requirements, financing requirements and projected financial performance, are forward-looking statements. In some cases, these forward-looking statements can be identified by words or phrases such as “may,” “will,” “expect,” “anticipate,” “aim,” “estimate,” “intend,” “plan,” “believe,” “potential,” “continue,” “is/are likely to” or other similar expressions. In addition, these forward-looking statements reflect our current views with respect to future events and are not a guarantee of future performance. Actual results may differ materially from the results predicted or implied by such statements, and Nebius’ reported results should not be considered as an indication of future performance. The potential risks and uncertainties that could cause actual results to differ from the results predicted or implied by such statements include, among others, our ability to successfully operate and develop a fundamentally different, early-stage group following the divestment of a significant portion of our historical operations; to successfully identify sites and enter into purchase, lease, build-to-suit or acquisition agreements with respect to additional data center capacity; to obtain cost-effective and reliable access to electrical power for such sites, to implement our business plans; to continue to successfully capture customers; to continue to successfully obtain required supplies of hardware on acceptable terms; and to obtain any further debt or equity financing that may be necessary to achieve our objectives. Many of these risks and uncertainties depend on the actions of third parties and are largely outside of our control. Notwithstanding the completion of the full divestment of our Russian businesses, we also continue to be subject to many of the risks and uncertainties included under the captions “Risk Factors” and “Operating and Financial Review and Prospects” in our Annual Report on Form 20-F for the year ended December 31, 2023 and “Risk Factors” in a shareholder circular filed as Exhibit 99.2 to a Report on Form 6-K filed with the
In addition, statements that “we believe” and similar statements reflect our beliefs and opinions on the relevant subject. These statements are based upon information available to us as of the date of this press release, and while we believe such information forms a reasonable basis for such statements, such information may be limited or incomplete, and our statements should not be read to indicate that we have conducted an exhaustive inquiry into, or review of, all potentially available relevant information. These statements are inherently uncertain, and investors are cautioned not to unduly rely upon these statements.
View source version on businesswire.com: https://www.businesswire.com/news/home/20250318096662/en/
For journalists: media@nebius.com
For investors: askIR@nebius.com
Source: Nebius Group
FAQ
When will Nebius (NBIS) launch NVIDIA Blackwell Ultra-powered instances?
What performance improvement does NVIDIA Dynamo offer on Nebius's platform?
Which U.S. data centers will host Nebius's NVIDIA Blackwell GPUs in 2025?
When will Nebius (NBIS) begin offering NVIDIA Blackwell GPU capacity in the US?
