Following a great Dell Technologies World event, we’re showing new server capabilities as part of the Dell AI Factory. These advancements are set to revolutionize the way businesses leverage artificial intelligence by providing a comprehensive, scalable and integrated approach to deploying AI solutions. Our new capabilities promise unprecedented AI performance power management and cooling, starting with the PowerEdge XE9680L with support for NVIDIA B200 HGX 8-way NVLink GPUs (graphics processing units). This offer pushes the boundaries of what’s possible for AI-driven operations by supporting up to 72 GPUs per rack at 107 kW and doubling I/O throughput.
Bringing AI to Your Data
It’s critical for customers to bring AI to their data to harness its full potential. But how can they achieve this more sustainably? The answer lies in deploying cutting-edge infrastructure that is optimized to handle the demands of AI workloads as efficiently as possible. Designed with Smart Power and Cooling, Dell PowerEdge servers and software are designed to help enable IT operations to maximize their power and thermal budgets.
Smart Power
By 2026, 50% of G20 members will experience monthly electricity rationing, turning energy-aware operations into either a competitive advantage or a major failure risk.¹ Dell PowerEdge servers seamlessly integrate robust hardware features with advanced software to provide a comprehensive Smart Power solution for data centers. The solution includes energy-efficient power supplies, optimized thermal designs and a suite of software that collectively reduces energy consumption and heat output while further refining power and cooling management.
Smart Cooling
Efficient power management is only one side of the equation. Remember, the power to cool is also critical. As we announced at Dell Technologies World 2024, Dell’s rack-scale solution with eight XE9680 H100 servers in a rack, with an integrated rear door heat exchanger, operates at 70 kW or less at the highest workloads. While Dell ensures component thermal and reliability requirements are met, we also innovate to minimize power needed to keep systems cool. Key hardware innovations such as rack-level integrated cooling, taller server chassis and the expansion of liquid cooling, including liquid-assisted air cooling (LAAC), work together to enhance heat dissipation, optimize airflow and support higher compute densities. One example for optimizing airflow is our efficient fan power management technology, which directly reduces operational costs by using an AI-based fuzzy logic controller for closed-loop thermal management.
Also, Dell’s Enterprise Infrastructure Planning Tool (EIPT), allows you to view power and greenhouse gas emissions as outputs of products, configurations and workloads.² Once servers are deployed, OpenManage Enterprise (OME) Power Manager provides monitoring and automation of server power and server telemetry including GPU data. Power Manager provides consumption reporting, thermal management and power capping policies. This increased visibility of information gives customers the ability to analyze and take action to maximize server power and cooling resources to deliver greater efficiencies.
Built for Reliability
Dell builds solutions with a clear focus on the data center and reliability. This is evident in our rigorous testing and validation processes to ensure that our systems can withstand the most demanding environments. A recent study highlighted issues with overheating in the data center, underscoring the importance of reliability in data center operations. In high-temperature test scenarios, a Dell PowerEdge HS5620 server continued running an intensive workload without component warnings or failures, while a Supermicro SYS‑621C-TN12R server failed.
Deployment and Professional Services
The Dell Rack Integration and MDC (Modular Data Center) teams provide extensive support for planning data center size and space allocation. It includes server and rack installation, airflow optimization and advanced power and cooling strategies to maximize efficiency and minimize environmental impact. Dell professionals aid in planning, implementing and managing sustainable data center operations, ensuring your AI infrastructure meets the demands of AI workloads. Dell Professional Services for Generative AI, guide you at every stage, from strategy alignment to implementation and support, ensuring secure optimized operations and accelerated AI outcomes.
Dell’s Ongoing Innovation
Dell has a long history of innovation focused on ensuring customers get the most energy efficiency³ out of recent technologies. For over 30 years, we’ve been at the forefront of developing server solutions that enhance power and energy efficiency. Our extensive portfolio of innovations related to power and thermal management is a testament to our dedication to this innovation. We’ve consistently introduced technologies that push the envelope, from the early days of power-efficient server designs to the latest AI-optimized solutions. Our commitment to delivering reliable and efficient systems has remained steadfast, ensuring that our customers always have access to the best technology and tools available.
We understand the critical role that reliable infrastructure plays in your operations. That’s why Dell PowerEdge servers are engineered to deliver consistent performance, even in the most challenging conditions. To learn more about our latest innovations and how they can benefit your business, visit Data Center Power and Cooling Solutions. We also invite you to join us at SC24 in November where you can see our solutions in action and speak with our experts.
1 Gartner’s 2024 predictions: Lots of AI, changing cybersecurity roles, electricity rationing, and more, Oct 18, 2023.
2 Dell’s Enterprise Infrastructure Planning Tool (EIPT) allows you to view power and Greenhouse Gas emissions as outputs of products, configurations, and workloads.
3 For more information on Energy Efficiency of Dell Products, see Energy Star – Active State Efficiency SERT Scores.
