HPE Ships First NVIDIA Grace Blackwell System

Hewlett Packard Enterprise (HPE) announced the shipment of its first NVIDIA Blackwell family-based solution, the NVIDIA GB200 NVL72. This rack-scale system is engineered to enable service providers and large enterprises to rapidly deploy extensive and complex Artificial Intelligence (AI) clusters. The system features advanced, direct liquid cooling solutions designed to optimize efficiency and performance.

The NVIDIA GB200 NVL72 stands out with its shared-memory, low-latency architecture, incorporating the latest GPU technology. It is designed to handle extremely large AI models exceeding a trillion parameters within a single memory space. This system seamlessly integrates NVIDIA CPUs, GPUs, compute and switch trays, networking, and software, delivering extreme performance for heavily parallelizable workloads like GenAI model training and inferencing, alongside NVIDIA software applications.

HPE’s extensive experience with liquid cooling, spanning five decades, positions the company to assist customers in achieving rapid deployment and robust infrastructure support for complex liquid-cooled environments. This expertise has enabled HPE to power eight of the top 15 supercomputers on the Green500 list, which ranks the world’s most energy-efficient supercomputers. HPE is recognized as a leader in direct liquid cooling technology.

Key features of the NVIDIA GB200 NVL72 by HPE include:

72 NVIDIA Blackwell GPUs and 36 NVIDIA Grace CPUs interconnected via high-speed NVIDIA NVLink
Up to 13.5 TB total HBM3e memory with 576 TB/sec bandwidth
HPE direct liquid cooling technology

HPE offers comprehensive services and support, enabling the delivery of AI solutions on a global scale. This includes supporting massive, custom AI clusters with superior serviceability, such as expert on-site support, customized services, and sustainability services. HPC & AI Custom Support Services are tailored to meet specific customer needs, offering enhanced incident management and proactive support through dedicated remote engineers.

Available services include:

Onsite engineering resources: Comprehensive on-site assistance is provided by highly skilled resident engineers who collaborate closely with a client’s IT departments to ensure peak system uptime and performance.
Performance and benchmarking engagements: Experts to fine-tune solutions throughout the system’s life.
Sustainability services: Resource monitoring, sustainability workshops, energy and emissions reporting.

The NVIDIA GB200 NVL72 is part of a wide array of high-performance computing and supercomputing systems by HPE designed to address various use cases for GenAI, scientific discovery, and other compute-intensive workloads.

FAQs

1. What is the NVIDIA GB200 NVL72?

The NVIDIA GB200 NVL72 is a rack-scale AI system developed by HPE, featuring 72 NVIDIA Blackwell GPUs and 36 NVIDIA Grace CPUs. It is designed for rapid deployment of large AI clusters, optimized with advanced direct liquid cooling technology.

2. Who is the target audience for this system?

The GB200 NVL72 is aimed at AI service providers and large enterprises that require scalable, high-performance computing solutions for complex AI workloads, including generative AI model training and inferencing.

3. How does the GB200 NVL72 position HPE in the market?

The launch of the GB200 NVL72 reinforces HPE’s leadership in AI infrastructure and high-performance computing, enabling the company to meet the growing demand for advanced AI solutions while leveraging its extensive experience in supercomputing and liquid cooling technologies.

Conclusion

The launch of the NVIDIA GB200 NVL72 by HPE represents a significant milestone in the realm of AI and high-performance computing. With its advanced architecture and powerful liquid cooling technology, this system is engineered to meet the demands of large enterprises and service providers for scalability and efficiency. The integration of NVIDIA’s cutting-edge GPUs and CPUs enables the handling of massive AI workloads, including trillion-parameter models, while maintaining energy efficiency. As organizations increasingly turn to AI solutions, the NVIDIA GB200 NVL72 positions HPE as a leader in delivering robust infrastructure capable of supporting the next generation of computational challenges. The collaboration between HPE and NVIDIA continues to foster advancements that will shape the future of AI computing.

To participate in our interviews, please write to our CyberTech Media Room at news@intentamplify.com

Tags: AI solutions, artificial intelligence, Grace Blackwell System, Hewlett Packard Enterprise, HPE, NVIDIA

CyberTech Staff Writer

CyberTech Staff Writer is a seasoned cybersecurity expert and analyst with a deep passion for safeguarding digital landscapes. With 20+ years of experience in the IT security and networking industry, they have honed their skills in identifying, assessing, and reporting cyber threats and best practices to salvage enterprises from disasters. Their expertise spans a wide range of areas, including cloud security, application security, ransomware assessment, threat intelligence, incident response, ZTNA and more. As a thought leader in the cybersecurity community, CyberTech Staff Writer is dedicated to sharing knowledge and insights to help organizations build robust defenses against evolving cyber threats.