As artificial intelligence (AI) workloads continue to grow, so too do the concerns surrounding energy consumption in data centers. Predictions indicate that energy use in these facilities may potentially double by 2030, a scenario that could exacerbate the pressure on existing power grids. Consequently, the need for reliable, resilient, and affordable energy solutions has become imperative for data center operations globally.
Innovating Data Center Efficiency
Dell Technologies is actively pursuing enhancements in efficiency across its AI solutions. This includes next-generation servers designed to reduce CPU power consumption by up to 65%. Along with advanced cooling technologies and improved software management systems, companies like Verne Global, Fulgent Genetics, and Emirates NBD have successfully reshaped their infrastructure and energy strategies, resulting in significant reductions in their ecological footprints.
Moving beyond infrastructure, Dell is exploring innovative methods to better match power demand with supply at the workload level. One of these initiatives is Concept Astro, which employs agentic AI, digital twins, and automation to optimize IT operations and energy usage. This concept envisions an adaptive AI Factory capable of forecasting task completion times and predicting energy requirements, using real-time grid data to determine the best time and place to execute various workloads while considering business priorities and energy costs.
Concept Astro’s Potential
Concept Astro offers a user-friendly dashboard for accessing real-time information and customized reports, making it easier for various roles—from researchers to CFOs—to manage energy-intensive workloads efficiently. This tool can inform decisions about when to run specific tasks, potentially reducing costs and emissions while maintaining business continuity.
Currently, Concept Astro is in the experimental phase, where digital twins are utilized to simulate workloads, enabling insights to enhance power usage effectively. The platform builds upon the Dell AIOps monitoring suite, providing users with tailored recommendations based on their system configurations.
Collaborative Success in Coral Reef Research
In collaboration with Scripps Institution of Oceanography at UC San Diego, Dell is testing Concept Astro to enhance coral reef research capabilities. By employing the Dell AI Factory, the team is converting extensive underwater imagery into high-resolution 3D models to assess coral health. Given the vast amount of data collected from dives each year, optimizing workload scheduling is critical to efficiency and reducing environmental impact.
A recent pilot program demonstrated how Concept Astro could help Scripps select optimal times to process workloads, leading to a 20% reduction in costs and a 32% decrease in emissions. Upgrading their data center equipment has also allowed the team to process twice the number of images in the same timeframe.
Advancing Energy Resilience
Dell is committed to helping businesses maximize their enterprise AI capabilities while minimizing energy consumption and associated operational costs. Continuous innovation and refinement of their technologies aim to deliver enhanced value to both businesses and society as a whole.
The company’s focus remains on assisting clients in overcoming significant challenges. With energy-efficient technologies transforming operations and pioneering concepts such as AI-driven workload scheduling, enterprises can see reduced costs, improved grid resilience, and notable progress in sustainability.
Further updates will be shared as Dell continues to explore strategies to enhance data center efficiency and grid awareness. In the meantime, insights from Dell’s Chief AI Officer John Roese and Chief Strategist Vivek Mohindra regarding AI, energy, and cooling are available for viewing online.