In an ambitious move towards shaping the future of data centers, Schneider Electric, a global leader in energy management and digital automation, has released an innovative guide. Dubbed ‘The AI Disruption: Challenges and Guidance for Data Center Design,’ (white paper 110), this latest publication is set to redefine the standard for AI-integrated data center infrastructure.
Given the meteoric rise of artificial intelligence in various sectors like healthcare, entertainment, manufacturing, and finance, data centers worldwide are grappling with new design and operational challenges. The need for strong processing power is growing exponentially as AI applications become essential to the business. This inexorably puts pressure on data centers to transform their strategy to meet these AI-driven computing demands.
Highlighting the gravity of this evolution, recent projections show AI workloads escalating at a CAGR of 26-36% by 2028. This rise predicates a parallel increase in energy requirements for both existing and upcoming data centers. Schneider Electric’s white paper 110, now available for public download, elaborates on vital considerations encompassing the four pillars of physical infrastructure: power, cooling, racks, and software tools.
AI Training Clusters
Schneider Electric's white paper emerges at a time when AI is reengineering business dynamics. The document would stand as a testament to the company's commitment to empowering businesses with data centers tailor-made for AI. Beyond a mere supportive role, the guide emphasizes creating an ecosystem that thrives on AI.
Pankaj Sharma, the Executive Vice President of the Secure Power Division and Data Center Business at Schneider Electric, remarked, “AI's rapid evolution necessitates a parallel rethinking of data center architecture. Especially with AI training clusters that demand massive processing prowess, the pressure on data center power and cooling systems is immense.” Sharma also underscored the mounting emphasis on green, high-efficiency systems in light of rising energy costs and heightened environmental awareness.
Evan Sparks, Chief Product Officer for Artificial Intelligence at Hewlett Packard Enterprise, lauded Schneider's initiative, asserting, “AI, poised to be an enterprise cornerstone, mandates a comprehensive stack design strategy. As AI solidifies its foothold in data centers, there's an escalating demand for AI-specific computational tools. Coupling this with proficient software platforms and sustainable data center solutions paves the way for effective and ethical AI deployment."
Power Densities, Liquid Cooling
Diving deep into AI's symbiosis with data center architecture, Schneider Electric's guide sheds light on pivotal aspects, including:
- Core AI Attributes - The guide offers a deep dive into the four fundamental AI trends affecting power, cooling, racks, and software management
- Rack Power Densities - Insights into accommodating the robust power demands of AI training servers are provided
- Transition to Liquid Cooling - As AI workloads surge, the manual recommends a switch from conventional air cooling to liquid cooling
- Rack Specifications - Detailed outlines to support AI servers, necessitating enhanced power, intricate cooling systems, and extensive networking cables
- Digital Twin Creation - The guide emphasizes the integration of DCIM, EPMS, and BMS software for a holistic digital replica of data center operations and assets
- Emerging Technologies - A forward-looking perspective on upcoming technologies and design methodologies to keep pace with AI's rapid evolution
In the rapidly advancing field of technology, Schneider Electric's most recent initiative exemplifies the company's dedication to developing ground-breaking solutions. This AI-focused handbook would serve as a testament to it foresight while setting a gold standard for contemporaries around the globe.
White Paper 110 is available for download here.
