HPE Announces Industry’s First 100% Fanless Direct Liquid Cooling Systems Architecture
HPE has unveiled the industry's first 100% fanless direct liquid cooling systems architecture to enhance energy and cost efficiency for large-scale AI deployments. This innovation was introduced at HPE's AI Day, showcasing the company's expertise in AI across various sectors. The new architecture aims to address the increasing power consumption in AI adoption by offering a more efficient cooling solution.
Key features of the system include:
- An 8-element cooling design
- High-density and high-performance system design
- Integrated network fabric design for massive scale
- Open system design for flexibility in accelerator choice
The architecture promises a 90% reduction in cooling power consumption compared to traditional air-cooled systems and a 37% reduction in cooling power required per server blade compared to hybrid direct liquid cooling. This innovation builds on HPE's experience in developing 7 of the top 10 supercomputers on the Green500 list.
HPE ha presentato l'architettura di sistemi di raffreddamento diretto a liquido completamente fanless della sua categoria per migliorare l'efficienza energetica e dei costi per le implementazioni su larga scala di AI. Questa innovazione è stata introdotta durante l'HPE AI Day, evidenziando l'esperienza dell'azienda nell'AI in vari settori. La nuova architettura mira a affrontare l'aumento del consumo energetico nell'adozione dell'AI offrendo una soluzione di raffreddamento più efficiente.
Le caratteristiche principali del sistema includono:
- Un design di raffreddamento a 8 elementi
- Design del sistema ad alta densità e alta prestazione
- Design della rete integrata per scala massiva
- Design del sistema aperto per flessibilità nella scelta degli acceleratori
L'architettura promette una riduzione del 90% nel consumo di energia per il raffreddamento rispetto ai sistemi tradizionali a raffreddamento ad aria e una riduzione del 37% nella potenza di raffreddamento necessaria per ogni blade server rispetto al raffreddamento diretto a liquido ibrido. Questa innovazione si basa sull'esperienza di HPE nello sviluppo di 7 dei primi 10 supercomputer nella lista Green500.
HPE ha presentado la primera arquitectura de sistemas de refrigeración líquida directa 100% sin ventiladores de la industria para mejorar la eficiencia energética y de costos en implementaciones a gran escala de IA. Esta innovación se introdujo en el Día de IA de HPE, mostrando la experiencia de la empresa en IA en diversos sectores. La nueva arquitectura tiene como objetivo abordar el aumento del consumo energético en la adopción de IA, ofreciendo una solución de refrigeración más eficiente.
Las características clave del sistema incluyen:
- Un diseño de refrigeración de 8 elementos
- Diseño del sistema de alta densidad y alto rendimiento
- Diseño de tejido de red integrado para una escala masiva
- Diseño de sistema abierto para flexibilidad en la elección de aceleradores
La arquitectura promete una reducción del 90% en el consumo de energía de refrigeración en comparación con los sistemas tradicionales enfriados por aire y una reducción del 37% en la energía de refrigeración requerida por cada blade server en comparación con la refrigeración líquida directa híbrida. Esta innovación se basa en la experiencia de HPE en el desarrollo de 7 de las 10 supercomputadoras más importantes de la lista Green500.
HPE는 대규모 AI 배포를 위한 에너지 및 비용 효율성을 향상시키기 위해 업계 최초의 100% 팬리스 직접 액체 냉각 시스템 아키텍처를 공개했습니다. 이 혁신은 HPE의 AI Day에서 소개되었으며, 다양한 분야에서 AI에 대한 회사의 전문성을 보여주었습니다. 새로운 아키텍처는 AI 채택에서 증가하는 전력 소비 문제를 다룰 수 있도록 보다 효율적인 냉각 솔루션을 제공합니다.
시스템의 주요 특징은 다음과 같습니다:
- 8 요소 냉각 디자인
- 고밀도 및 고성능 시스템 디자인
- 대규모를 위한 통합 네트워크 패브릭 디자인
- 가속기 선택의 유연성을 위한 오픈 시스템 디자인
이 아키텍처는 전통적인 공기 냉각 시스템에 비해 냉각 전력 소비를 90% 줄이는 것을 약속하며, 하이브리드 직접 액체 냉각에 비해 서버 블레이드당 필요한 냉각 전력을 37% 줄이는 것을 약속합니다. 이 혁신은 HPE가 Green500 목록의 상위 10개 슈퍼컴퓨터를 개발한 경험을 바탕으로 합니다.
HPE a dévoilé l'architecture des systèmes de refroidissement direct liquide 100% sans ventilateur de l'industrie pour améliorer l'efficacité énergétique et de coût pour les déploiements d'IA à grande échelle. Cette innovation a été présentée lors de la Journée de l'IA de HPE, mettant en avant l'expertise de l'entreprise en matière d'IA dans divers secteurs. La nouvelle architecture vise à répondre à l'augmentation de la consommation d'énergie dans l'adoption de l'IA en offrant une solution de refroidissement plus efficace.
Les caractéristiques clés du système comprennent :
- Un design de refroidissement à 8 éléments
- Un design de système à haute densité et haute performance
- Un design de réseau intégré pour une échelle massive
- Un design de système ouvert pour plus de flexibilité dans le choix des accélérateurs
L'architecture promet une réduction de 90% de la consommation d'énergie pour le refroidissement par rapport aux systèmes traditionnels refroidis par air et une réduction de 37% de l'énergie de refroidissement requise par bloc serveur par rapport au refroidissement direct liquide hybride. Cette innovation s'appuie sur l'expérience de HPE dans le développement de 7 des 10 superordinateurs les plus performants de la liste Green500.
HPE hat die erste 100% lüfterlose Architektur von Direkttemperaturregelungssystemen in der Branche vorgestellt, um die Energie- und Kosteneffizienz von großangelegten KI-Einsätzen zu verbessern. Diese Innovation wurde am HPE AI Day präsentiert und zeigt die Expertise des Unternehmens im Bereich KI in verschiedenen Sektoren. Die neue Architektur zielt darauf ab, dem zunehmenden Energieverbrauch bei der KI-Nutzung entgegenzuwirken, indem sie eine effizientere Kühlungslösung bietet.
Zu den wichtigsten Funktionen des Systems gehören:
- Ein 8-Elemente-Kühlungsdesign
- Hochdichtes und hochleistungsfähiges Systemdesign
- Integriertes Netzwerkdesign für massive Skalierung
- Offenes Systemdesign für Flexibilität bei der Auswahl von Beschleunigern
Die Architektur verspricht eine 90%ige Reduzierung des Kühlenergieverbrauchs im Vergleich zu herkömmlichen luftgekühlten Systemen und eine 37%ige Reduzierung der erforderlichen Kühlleistung pro Serverblade im Vergleich zur hybriden Direkttemperaturregelung. Diese Innovation basiert auf HPEs Erfahrung in der Entwicklung von 7 der besten 10 Supercomputer auf der Green500-Liste.
- Introduction of industry's first 100% fanless direct liquid cooling systems architecture
- Potential 90% reduction in cooling power consumption compared to traditional air-cooled systems
- 37% reduction in cooling power required per server blade compared to hybrid direct liquid cooling
- Reduced utility costs, carbon production, and data center fan noise
- Increased server cabinet density, resulting in less floor space consumption
- None.
Insights
HPE hosts AI Day at state-of-the-art manufacturing facility, showcases decades of leadership in developing direct liquid cooled AI systems at scale
Industry’s first
While efficiency has improved in next-generation accelerators, power consumption is continuing to intensify with AI adoption, outstripping traditional cooling techniques.
Organizations running large AI workloads will need to do so more efficiently. The most effective way to cool next-generation AI systems is through direct liquid cooling, of which HPE is a pioneer. This critical cooling technology has enabled HPE’s systems to deliver 7 of the top 10 supercomputers on the Green500 list, which ranks the world’s most energy-efficient supercomputers.
Based on this expertise, HPE’s
“As organizations embrace the possibilities created by generative AI, they also must advance sustainability goals, combat escalating power requirements, and lower operational costs,” said Antonio Neri, president and CEO of HPE. “The architecture we unveiled today uses only liquid cooling, delivering greater energy and cost-efficiency advantages than the alternative solutions on the market. In fact, this direct liquid cooling architecture has the potential to yield a
The system architecture is built on four pillars:
- An 8-element cooling design that includes liquid cooling for the GPU, CPU, full server blade, local storage, network fabric, rack/cabinet, pod/cluster and coolant distribution unit (CDU)
- High-density and high-performance system design, complete with rigorous testing, monitoring software, and on-site services to support successful deployment of these sophisticated compute and cooling systems
- Integrated network fabric design for massive scale integrating lower-cost and lower-power connections
- Open system design to offer flexibility of choice in accelerators
The
HPE’s leadership and unique market opportunity
At AI Day, Antonio Neri, president and CEO; Fidelma Russo, EVP & GM, Hybrid Cloud and HPE CTO; and Neil MacDonald, EVP & GM, Server; discussed how the HPE portfolio comprises the critical building blocks of networking, storage and hybrid cloud to deliver on the promise of AI.
Read about Neri’s reflections on HPE’s AI Day in his blog post.
About Hewlett Packard Enterprise
Hewlett Packard Enterprise (NYSE: HPE) is the global edge-to-cloud company that helps organizations accelerate outcomes by unlocking value from all of their data, everywhere. Built on decades of reimagining the future and innovating to advance the way people live and work, HPE delivers unique, open and intelligent technology solutions as a service. With offerings spanning Cloud Services, Compute, High Performance Computing & AI, Intelligent Edge, Software, and Storage, HPE provides a consistent experience across all clouds and edges, helping customers develop new business models, engage in new ways, and increase operational performance. For more information, visit: www.hpe.com.
Forward-looking statements
This press release contains forward-looking statements within the meaning of the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. Such statements involve risks, uncertainties, and assumptions. If the risks or uncertainties ever materialize or the assumptions prove incorrect, the results of Hewlett Packard Enterprise Company and its consolidated subsidiaries may differ materially from those expressed or implied by such forward-looking statements and assumptions. The words “believe,” “expect,” “anticipate,” “guidance,” “intend,” “will,” "may," and similar expressions are intended to identify such forward-looking statements. All statements other than statements of historical fact are statements that could be deemed forward-looking statements, including but not limited to any projections, estimations, or expectations of addressable markets and their sizes, operational or financial performance, plans, strategies and objectives of management for future operations or performance. Risks, uncertainties and assumptions include those that are described in HPE’s Annual Report on Form 10-K for the fiscal year ended October 31, 2023, Quarterly Reports on Form 10-Q, Current Reports on Form 8-K, and that are otherwise described or updated from time to time in Hewlett Packard Enterprise’s Securities and Exchange Commission reports. Hewlett Packard Enterprise assumes no obligation and does not intend to update these forward-looking statements, except as required by applicable law.
View source version on businesswire.com: https://www.businesswire.com/news/home/20241010585008/en/
Media Contact:
Jen Colton
jen.colton@hpe.com
Source: Hewlett Packard Enterprise
FAQ
What new cooling technology did HPE announce for AI systems?
How much cooling power can be saved with HPE's new liquid cooling system?
What are the key features of HPE's new cooling system architecture?
How does HPE's new cooling system impact server blade efficiency?
Where did HPE showcase its new AI cooling technology (HPE stock symbol: HPE)?
