Block Becomes First Company in North America to Deploy the Latest NVIDIA GB200 Systems for Frontier Models
Block (NYSE: XYZ) has announced its position as North America's first company to deploy the NVIDIA DGX SuperPOD with DGX GB200 systems. The infrastructure will be installed at an Equinix data center, focusing on research and training of open source generative AI models.
The company's CTO, Dhanji R. Prasanna, emphasized Block's commitment to applying AI solutions and advancing frontier AI technology through an open source approach. Block's previous AI research includes work on deepfake detection and hyper-realistic generated audio.
The deployment leverages Lambda 1-Click Clusters for initial testing, providing access to hundreds of interconnected NVIDIA GPUs. The infrastructure will be housed in Equinix's AI-ready data centers, offering data privacy, high-performance computing, and low-latency connectivity. This announcement follows Block's recent launch of 'codename goose,' an open source AI agent framework connecting large language models to real-world actions.
Block (NYSE: XYZ) ha annunciato la sua posizione come prima azienda del Nord America a implementare il NVIDIA DGX SuperPOD con sistemi DGX GB200. L'infrastruttura sarà installata in un centro dati Equinix, concentrandosi sulla ricerca e formazione di modelli di intelligenza artificiale generativa open source.
Il CTO dell'azienda, Dhanji R. Prasanna, ha sottolineato l'impegno di Block nell'applicare soluzioni di intelligenza artificiale e nel promuovere la tecnologia AI all'avanguardia attraverso un approccio open source. Le precedenti ricerche di AI di Block includono lavori sulla rilevazione di deepfake e audio generato iper-realistico.
Il dispiegamento sfrutta i Cluster Lambda 1-Click per i test iniziali, fornendo accesso a centinaia di GPU NVIDIA interconnesse. L'infrastruttura sarà ospitata nei centri dati pronti per l'AI di Equinix, offrendo privacy dei dati, calcolo ad alte prestazioni e connettività a bassa latenza. Questo annuncio segue il recente lancio da parte di Block di 'codename goose', un framework di agente AI open source che collega modelli di linguaggio di grandi dimensioni ad azioni nel mondo reale.
Block (NYSE: XYZ) ha anunciado su posición como la primera empresa de América del Norte en desplegar el NVIDIA DGX SuperPOD con sistemas DGX GB200. La infraestructura se instalará en un centro de datos de Equinix, centrando su atención en la investigación y formación de modelos de IA generativa de código abierto.
El CTO de la empresa, Dhanji R. Prasanna, enfatizó el compromiso de Block con la aplicación de soluciones de IA y el avance de la tecnología de IA de frontera a través de un enfoque de código abierto. Las investigaciones previas de IA de Block incluyen trabajos en la detección de deepfakes y audio generado hiperrealista.
El despliegue aprovecha los Clusters Lambda 1-Click para pruebas iniciales, proporcionando acceso a cientos de GPU NVIDIA interconectadas. La infraestructura se alojará en los centros de datos listos para IA de Equinix, ofreciendo privacidad de datos, computación de alto rendimiento y conectividad de baja latencia. Este anuncio sigue al reciente lanzamiento de Block de 'codename goose', un marco de agente de IA de código abierto que conecta modelos de lenguaje grandes con acciones del mundo real.
Block (NYSE: XYZ)는 북미 최초로 NVIDIA DGX SuperPOD와 DGX GB200 시스템을 배포한다고 발표했습니다. 이 인프라는 Equinix 데이터 센터에 설치되며, 오픈 소스 생성 AI 모델의 연구 및 교육에 중점을 둡니다.
회사의 CTO인 Dhanji R. Prasanna는 Block이 AI 솔루션을 적용하고 오픈 소스 접근 방식을 통해 첨단 AI 기술을 발전시키겠다는 의지를 강조했습니다. Block의 이전 AI 연구에는 딥페이크 탐지 및 초현실적으로 생성된 오디오에 대한 작업이 포함됩니다.
배포는 초기 테스트를 위해 Lambda 1-Click 클러스터를 활용하여 수백 개의 상호 연결된 NVIDIA GPU에 대한 접근을 제공합니다. 이 인프라는 Equinix의 AI 준비 데이터 센터에 수용되어 데이터 프라이버시, 고성능 컴퓨팅 및 저지연 연결성을 제공합니다. 이 발표는 Block이 최근에 출시한 'codename goose'의 발표에 이어지는 것으로, 이는 대형 언어 모델을 실제 행동과 연결하는 오픈 소스 AI 에이전트 프레임워크입니다.
Block (NYSE: XYZ) a annoncé sa position en tant que première entreprise d'Amérique du Nord à déployer le NVIDIA DGX SuperPOD avec des systèmes DGX GB200. L'infrastructure sera installée dans un centre de données Equinix, axée sur la recherche et la formation de modèles d'IA générative open source.
Le CTO de l'entreprise, Dhanji R. Prasanna, a souligné l'engagement de Block à appliquer des solutions d'IA et à faire progresser la technologie de l'IA de pointe grâce à une approche open source. Les recherches précédentes en IA de Block comprennent des travaux sur la détection des deepfakes et l'audio généré hyperréaliste.
Le déploiement utilise les Clusters Lambda 1-Click pour des tests initiaux, offrant un accès à des centaines de GPU NVIDIA interconnectés. L'infrastructure sera hébergée dans les centres de données prêts pour l'IA d'Equinix, offrant confidentialité des données, calcul haute performance et connectivité à faible latence. Cette annonce fait suite au récent lancement par Block de 'codename goose', un cadre d'agent IA open source reliant de grands modèles de langage à des actions du monde réel.
Block (NYSE: XYZ) hat seine Position als erstes Unternehmen in Nordamerika bekannt gegeben, das das NVIDIA DGX SuperPOD mit DGX GB200-Systemen einführt. Die Infrastruktur wird in einem Equinix-Rechenzentrum installiert und konzentriert sich auf die Forschung und Schulung von Open-Source-Modellen für generative KI.
Der CTO des Unternehmens, Dhanji R. Prasanna, betonte das Engagement von Block, KI-Lösungen anzuwenden und die Technologie der Grenz-KI durch einen Open-Source-Ansatz voranzutreiben. Die bisherigen KI-Forschungen von Block umfassen Arbeiten zur Erkennung von Deepfakes und hyperrealistisch generiertem Audio.
Der Einsatz nutzt Lambda 1-Click-Cluster für erste Tests und bietet Zugang zu Hunderten von miteinander verbundenen NVIDIA-GPUs. Die Infrastruktur wird in den KI-bereiten Rechenzentren von Equinix untergebracht, die Datenschutz, Hochleistungsrechnen und latenzarme Konnektivität bieten. Diese Ankündigung folgt auf Blocks kürzlichen Start von 'codename goose', einem Open-Source-KI-Agenten-Framework, das große Sprachmodelle mit realen Aktionen verbindet.
- First company in North America to deploy cutting-edge NVIDIA DGX SuperPOD with GB200 systems
- Strategic partnership with Equinix for advanced AI infrastructure deployment
- Demonstrated AI capabilities with successful projects in deepfake detection and audio generation
- Recent launch of 'codename goose' shows practical implementation of AI technology
- Significant capital investment required for high-end AI infrastructure
- Competitive pressure to maintain technological leadership in AI development
Insights
Block's deployment of NVIDIA's GB200 systems represents a significant strategic investment in frontier AI capabilities that positions the company at the technological forefront. While the announcement lacks specific investment figures, securing first-mover status for this cutting-edge infrastructure provides Block with potential competitive advantages in developing proprietary AI applications.
The partnership structure with NVIDIA and Equinix creates a powerful technology ecosystem that combines Block's financial technology expertise with state-of-the-art AI hardware and global data center infrastructure. This three-way collaboration enables Block to leverage specialized AI capabilities while maintaining the flexibility of Equinix's cloud-adjacent platform.
Notably, Block's commitment to an open source approach is strategically significant. This model can accelerate innovation through community contributions while potentially reducing R&D costs compared to closed development systems. The focus on "underexplored areas" suggests Block is targeting AI applications with competition, potentially creating new revenue streams or enhancing existing product offerings.
Investors should recognize that while this infrastructure investment demonstrates technological leadership, concrete financial benefits will depend on Block's ability to translate these AI capabilities into practical applications that either generate revenue or improve operational efficiency for its payment ecosystem.
Block's acquisition of NVIDIA's GB200 systems positions them among the elite tier of companies with access to the most advanced AI hardware available. The DGX SuperPOD architecture represents purpose-built infrastructure for frontier model development - the large-scale generative AI systems that define competitive advantage in today's AI landscape.
Their technical approach reflects sophisticated AI strategy: they've intelligently staged their infrastructure journey by first utilizing Lambda 1-Click Clusters for hypothesis testing before committing to permanent infrastructure. This methodology demonstrates disciplined technology investment practices that mitigate risk.
Block's previous work in deepfake detection and hyper-realistic audio generation provides a foundation of relevant AI expertise that increases the probability of successful implementation with this new hardware. Their "codename goose" framework for connecting LLMs to real-world actions points toward practical applications rather than purely research-focused activities.
The deployment at Equinix's AI-ready facilities adds further technical advantages through low-latency connectivity to cloud providers and edge capabilities. This architecture provides flexibility for future scaling while maintaining data sovereignty - a critical consideration for financial technology applications where privacy and regulatory compliance are paramount.
This infrastructure positions Block to potentially develop proprietary AI capabilities that could differentiate their financial products or create entirely new business models in the intersection of AI and financial services.
Block’s AI research will focus on solving problems to empower real people in the economy using open source technology
DISTRIBUTED-WORK-MODEL/
“The industry, and the world, is undergoing a seismic shift with adopting AI tools. At Block, we think it’s essential not only to apply AI to existing problems, but also to explore, learn, and build in the open so that we can advance the frontier of AI in a way that truly levels the playing field for our customers and community,” said Dhanji R. Prasanna, CTO of Block. “We’re excited to deploy an NVIDIA Grace Blackwell DGX SuperPOD and start exploring novel solutions for our customers. We're committed to an open source approach, sharing our learnings and results along the way.”
Block’s AI research has previously worked on deepfake detection, and attracted widespread attention for one of the first examples in the world of hyper-realistic generated audio.
“Block is a technology company first, and engineering is at the heart of everything we do. Our past work on generative speech models means Block is well-positioned for innovation surrounding Gen AI models, and all possible applications that come with it,” said Prasanna.
The Grace Blackwell-powered NVIDIA DGX SuperPOD is purpose-built for state-of-the-art generative AI models, bringing advanced capabilities for training and inference. These systems address key AI challenges, enabling larger and more advanced models to be trained, with faster innovation and greater capabilities for AI as a result.
“As AI models grow in complexity and scale, businesses need powerful infrastructure that can match the pace of innovation,” said Charlie Boyle, Vice President, DGX platforms, NVIDIA. “With NVIDIA DGX GB200 systems, Block engineering and research teams can develop frontier open source AI models that can tackle complex, real-world challenges with state-of-the-art AI supercomputing.”
A key to select the right AI Cloud partner for Block and prove out hypotheses before scaling was the ability to access hundreds of interconnected NVIDIA GPUs, for a short amount of time and supported by ML engineers: Lambda 1-Click Clusters. These GPU clusters are now available with NVIDIA Blackwell.
The new DGX SuperPOD will be deployed at one of Equinix’s AI-ready data centers. These purpose-built, globally interconnected facilities provide a unique environment of data privacy and sovereignty compliance, high-performance, flexibility and low-latency edge connectivity to thousands of ecosystem partners and clouds.
“Frontier models represent the cutting edge of artificial intelligence technology, pushing the boundaries of what AI can achieve, and they require the latest in AI chips – like NVIDIA’s new DGX SuperPOD,” said Jon Lin, Chief Business Officer at Equinix. “By deploying at Equinix’s neutral, cloud-adjacent platform, companies like Block can unlock expanded compute scale and flexibility. This enables the customization of AI solutions with a choice of infrastructure, cloud, models and cooling at our neutral exchange.”
Adopting NVIDIA’s latest AI infrastructure is a fast follow from Block’s launch of codename goose, an open source, interoperable AI agent framework that enables users to connect large language models (LLMs) to real-world actions. Its first use cases are related to software engineering, but developers both within Block and the broader open source community are exploring other non-engineering use cases as well.
About Block
Block, Inc. (NYSE: XYZ) builds technology to increase access to the global economy. Each of our brands unlocks different aspects of the economy for more people. Square makes commerce and financial services accessible to sellers. Cash App is the easy way to spend, send, and store money. Afterpay is transforming the way customers manage their spending over time. TIDAL is a music platform that empowers artists to thrive as entrepreneurs. Bitkey is a simple self-custody wallet built for bitcoin. Proto is a suite of bitcoin mining products and services. Together, we’re helping build a financial system that is open to everyone. Block.xyz
View source version on businesswire.com: https://www.businesswire.com/news/home/20250312684830/en/
Media Contact
press@block.xyz
Investor Relations Contact
ir@block.xyz
Source: Block, Inc.
FAQ
What is the significance of Block (XYZ) being the first to deploy NVIDIA GB200 Systems in North America?
How will Block's NVIDIA GB200 deployment impact its AI development capabilities?
Where will Block (XYZ) deploy its new NVIDIA DGX SuperPOD system?
What are Block's (XYZ) current AI initiatives and developments?
