Skip to main content
Arrow Electronics, Inc.

Arrow Quick Hit: NVIDIA DGX H100 BasePOD

April 24, 2024 | Russ Braden

The infrastructure foundation for enterprise AI

What is it?

The number of opportunities to utilize AI in enterprise businesses and other organizations continues to expand quickly. AI infrastructure demands significant compute resources for efficient model training, often across distributed clusters. While cloud computing offers a straightforward approach, many enterprises prefer on-premises data centers, despite the complexities involved in building AI infrastructure locally. Careful planning and coordination will make the deployment and day-to-day operations easier for system administrators. 

NVIDIA DGX H100 BasePOD provides the essential infrastructure and software needed to accelerate the deployment and execution of AI workloads. It leverages the success of NVIDIA DGX Systems to eliminate the design challenges and management complexities associated with scaling AI infrastructure and provides an optimized foundation for AI development. 

How does it work?

The DGX H100 BasePOD combines NVIDIA's high-performance computing hardware with optimized software and networking solutions to provide a turnkey AI infrastructure solution. 


NVIDIA DGX H100 BasePOD core hardware components

AI compute

The NVIDIA DGX H100 System is the compute engine for BasePOD jobs. This system is designed to maximize AI throughput and provides enterprises with a highly refined, systematic and scalable platform to help achieve breakthroughs in natural language processing, recommender systems, data analytics and more. Featuring the NVIDIA H100 Tensor Core GPU, the DGX H100 works to consolidate training, inference and analytics into a unified, easy-to-deploy AI infrastructure that includes direct access to NVIDIA AI experts. 

NVIDIA Base Command powers every DGX BasePOD. It offers Kubernetes, Slurm and Jupyter Notebook environments for DGX Systems and ensures enterprise-grade scheduling and orchestration. With unified management and NVIDIA support for meeting SLAs, it automates DGX system management from single nodes to large scale clusters. 

Click on image to view a larger version.

NVIDIA Enterprise AI and data analytics

AI software

NVIDIA AI Enterprise encompasses a collection of AI and data analytics software for AI development and deployment. It incorporates trusted open-source containers and frameworks like NVIDIA RAPIDS, NVIDIA TAO Toolkit, NVIDIA TensorRT and NVIDIA Triton Inference Server, all certified and supported for DGX systems. NVIDIA AI Enterprise complements NVIDIA Base Command for comprehensive AI utilization and is included with DGX systems. 


NVIDIA QH Infiniband

NVIDIA DGX H100 BasePOD core hardware components


InfiniBand and Ethernet technologies enable networking functionality in the DGX H100 BasePOD that provides maximum bandwidth for the GPUs to communicate. 


The NVIDIA SN5600 switch offers a total of 256x 200GbE, 128x 400GbE ports for in-band management of the DGX BasePOD. The SN5600 switch is used for Ethernet connectivity for storage appliances as well. The ports on the NVIDIA DGX are shared with the in-band management Ethernet ports. 


NVIDIA SN2201 switches have 48 ports that that connect out of band (OOB) management and consolidate management connectivity for all components in BasePOD.

Below is a breakdown of the minimum required hardware technology components that make up the NVIDIA DGX H100 BasePOD.

Click on image to view a larger version.

Centralized storage array

Multiple, certified partner vendor flash storage arrays

Offered as fully integrated and deployable solutions via the NVIDIA Partner Network, DGX BasePOD configurations vary in scale, spanning from two to hundreds of DGX systems. These configurations feature certified high-performance storage solutions from NVIDIA DGX Storage Technology Partners such as DDN, Dell, NetApp, Pure Storage, VAST Data and WEKA.

Why should you care?

According to IDC, worldwide spending on AI-centric systems will pass $300 billion by 2026. The high and consistent growth seen in the use of AI in all industries demonstrates its importance to future business. "AI is not the future, it is now," said Mike Glennon, senior market research analyst with IDC's Customer Insights and Analysis team. "Most IT vendors have adopted AI solutions to supplement their products and are enhancing their products to make AI crucial to their success. Those vendors that are only now considering AI are at a considerable disadvantage to IT vendors that have AI-based products already in production and AI is becoming crucial to the capabilities of many products."

Now is the time to capitalize on the growing demand for AI infrastructure solutions by selling the NVIDIA DGX H100 BasePOD. Here are some reasons to engage with customers now:

  • High demand: Meet customers' need for cutting-edge AI computing capabilities. 
  • Brand recognition: NVIDIA is a well-known and trusted brand in the AI and computing industries. 
  • Profitable: The DGX H100 BasePOD may come with higher margins compared to commodity hardware or software. 
  • Comprehensive support: NVIDIA provides marketing materials, technical assistance and sales enablement resources for the DGX H100 BasePOD.
  • Long-term partnerships: Customers may require ongoing support, upgrades and additional services related to their AI infrastructure. 

Differentiation in the market

These easy-to-sell turnkey solutions offer the performance and scale that your customers require. They are pre-verified, pre-tested and installed by Arrow so customers can start putting AI to work faster than ever before. 

Key benefits of the DGX H100 BasePOD

  • Simplified deployment: This turnkey solution minimizes the need for extensive setup and configuration, significantly reducing the complexity and time investment required to get AI systems up and running efficiently. 
  • Scalability: With its modular design, customers can start small and expand seamlessly as their requirements grow. 
  • Reliability and performance: BasePOD processes complex AI tasks efficiently and effectively. 
  • Optimized storage technology: BasePOD ensures compatibility and expected performance for various workloads, supported by rigorous testing from NVIDIA's validated storage partners.
  • Support and maintenance: BasePOD comes with full support including troubleshooting, upgrades and maintenance to minimize downtime and keep AI systems running smoothly. 

How do you position and sell the NVIDIA DGX BasePOD?

DGX BasePOD is the go-to solution for organizations looking to deploy AI workloads at scale. Target the following personas for best results:

  • Data scientists/AI researchers who require high-performance computing systems for training and experimentation with deep learning models
  • AI developers responsible for building AI applications and solutions
  • AI engineers responsible for deploying AI solutions in production environments
  • IT infrastructure managers responsible for evaluating and deploying new infrastructure solutions within the organization
  • Executives who prioritize innovation and seek competitive advantages through AI adoption
  • Procurement managers who make purchasing decisions and negotiate contracts with vendors

Essential questions to understand AI needs

  • What specific AI workloads or projects are you looking to accelerate?
  • What are your current pain points with deploying and managing AI infrastructure?
  • What level of scalability do you require for your AI workloads?
  • What are your performance and reliability expectations for AI computing infrastructure?
  • Are there any regulatory or compliance requirements that need to be considered in your AI deployments?
  • How important are AI support and maintenance services to your organization?
  • What specific integrations or compatibility requirements with other software or tools will you need?
  • What is your budget for investing in AI infrastructure, including hardware, software and ongoing support?
  • What are your long-term goals and objectives for AI adoption?

Questions you may be asked

  • How does the DGX H100 BasePOD compare to previous DGX systems in terms of performance metrics?
  • What options are available for scaling up the DGX H100 BasePOD to larger configurations?
  • Is the DGX H100 Software Stack customizable?
  • What considerations should be made when deploying the DGX H100 BasePOD in terms of power, cooling and networking?
  • What are some typical use cases in which the DGX H100 BasePOD excels?
  • How does the Total Cost of Ownership (TCO) of the DGX H100 BasePOD compare to alternative solutions?
  • What support options are available for the DGX H100 BasePOD?
  • Is there a dedicated support team?

Contact the Arrow NVIDIA team to help you drive your DGX H100 BasePOD sales today!

Learn more about NVIDIA solutions available through Arrow. 

Subscribe to Arrow Channel Advisor
Sign Up