ABB and NVIDIA have announced a strategic partnership to design next-generation electrical infrastructure capable of powering gigawatt-scale AI data centers. The collaboration focuses on developing high-efficiency power distribution systems using NVIDIA's proposed 800 VDC (volt direct current) framework, which will enable 1-megawatt server racks to operate at unprecedented efficiency levels.

Why Are Data Centers Struggling With Power Delivery?

The explosive growth of artificial intelligence has created an unprecedented challenge: delivering massive amounts of electrical power to densely packed computing equipment. Traditional power systems, designed decades ago, simply cannot handle the scale of modern AI workloads. As AI adoption accelerates globally, data centers need fundamentally new approaches to distribute power efficiently and reliably. The partnership between ABB, a global leader in electrical engineering, and NVIDIA, the dominant maker of graphics processing units (GPUs) used in AI systems, represents an attempt to solve this infrastructure bottleneck before it becomes a critical constraint on AI development.

The scale of the challenge is staggering. Industry projections indicate global data center capacity will surge from 80 gigawatts in 2024 to approximately 220 gigawatts by 2030, with infrastructure investments expected to exceed $1 trillion. Roughly 70% of this expansion will be driven by AI-specific workloads.

What Technologies Will Power These New Data Centers?

The ABB-NVIDIA partnership will focus on implementing several cutting-edge electrical technologies designed to handle extreme power densities. The approach centers on integrating medium voltage (MV) uninterruptible power supply (UPS) units, which provide backup power during outages, with direct current (DC) distribution networks that use solid-state power electronics. These solid-state components, which have no moving parts, offer superior efficiency and faster response times compared to traditional electromechanical systems.

ABB has already developed innovative products supporting this vision. The company created HiPerGuard, described as the world's first solid-state MV UPS, which enhances power density and energy efficiency in compact installations. ABB also introduced the SACE Infinitus, the industry's first IEC-certified solid-state circuit breaker, which provides ultra-fast response times and precise control essential for effective direct current distribution.

How Are Companies Preparing for AI's Power Demands?

• Advanced Power Management: ABB's data center solutions feature sophisticated power management systems that monitor and optimize energy consumption in real time, ensuring AI servers operate at peak efficiency while minimizing waste.
• Emergency Energy Reserves: Uninterruptible power supply systems provide backup power during grid disruptions, protecting expensive AI hardware and ensuring continuous operation of critical workloads.
• Digital Performance Analytics: Real-time monitoring and analytics help data center operators identify inefficiencies and optimize their electrical infrastructure, reducing both costs and environmental impact.
• Thermal Management Systems: Cooling represents a major portion of data center energy consumption; advanced thermal management technologies help dissipate the intense heat generated by densely packed GPUs.
• Protective Equipment: Solid-state circuit breakers and other protective devices safeguard equipment from electrical faults and surges, preventing costly damage and downtime.

The commitment to this technology shift is substantial. Approximately 40% of ABB's electrification research and development focuses on technologies vital for next-generation data center facilities, including electrical frameworks, protective equipment, DC networks, and thermal management systems.

"ABB is spearheading the development of critical electrical distribution technologies shaping tomorrow's data centers. Our early investments in advanced UPS systems, DC infrastructure, and solid-state components position us to address AI's escalating energy needs. Together with NVIDIA, we're driving industry adoption of 800 volt architectures essential for supporting next-level AI infrastructure," stated Giampiero Frisio, President of ABB Electrification.

Giampiero Frisio, President of ABB Electrification

NVIDIA's perspective emphasizes the practical benefits of this collaboration. The company recognizes that as AI adoption accelerates, data centers require innovative power distribution strategies that enhance efficiency and streamline system designs. The 800-volt architecture being developed will underpin the high-density computing environments necessary for future AI breakthroughs.

"As global AI adoption accelerates, data centers require innovative power distribution strategies that enhance efficiency and streamline system designs. This joint effort with ABB is advancing 800 volt architectures that will underpin the high-density computing environments necessary for future AI breakthroughs," noted Dion Harris, senior director of HPC, Cloud and AI Infrastructure at NVIDIA.

Dion Harris, Senior Director of HPC, Cloud and AI Infrastructure at NVIDIA

The partnership signals a broader industry recognition that power infrastructure, not computing chips themselves, may become the limiting factor in AI expansion. By collaborating on electrical systems now, ABB and NVIDIA are attempting to ensure that the physical infrastructure can keep pace with the exponential growth in AI computing demand. This work will likely influence how data centers are designed and built for the next decade, making it one of the most consequential engineering efforts in the AI industry.