As we look ahead to 2026, the core battleground in artificial intelligence infrastructure is shifting from the endless chase for more GPUs to the race for faster and more efficient electricity supply. Pioneered by NVIDIA Corporation(NVDA.US), the industry is moving toward high-voltage direct current (HVDC) solutions, with its breakthrough 800V HVDC architecture set to redefine data center power standards.

From Bottlenecks to Breakthroughs: The New Power Paradigm

In 2025, Nvidia rolled out its 800V HVDC blueprint, leveraging solid-state transformers to convert 10-20kV AC directly to 800V DC at the data center edge. According to VERTIV HOLDINGS LLC(VRT.US), this revolution is as radical as upgrading country lanes into highways: doubling the voltage cuts the current in half, with major efficiency and space-saving benefits.

Key advantages of 800V HVDC:

Higher efficiency: Lower current means less heat, streamlining energy flow from the grid to the chips.
Space savings: Smaller current supports reduced copper and busbar usage, letting facilities scale up flexibly.
Simplified power path: Fewer conversion steps mean faster, cleaner energy transmission.

Goldman Sachs forecasts the addressable market for AI data center-related power products will expand at a compound annual growth rate (CAGR) of 39% from 2025 to 2030—including gas turbines, transformers, UPS systems, server power supplies, electrical components, and liquid cooling.

Three growth pillars stand out: ongoing data center capacity expansion, rising rack power density, and a major shift from AC to DC architectures. With AI rack power requirements accelerating from 100kW toward 1MW, HVDC is no longer just a trend—it’s becoming the next industry standard. According to Goldman Sachs, utilizing higher-voltage DC can reduce facility energy consumption by 5–15%, and Nvidia, along with its ecosystem partners, is fully committed to this transition. As cloud service providers (CSPs) are the ultimate buyers, companies that supply HVDC hardware and systems are increasingly making it onto supplier shortlists for next-generation AI deployments.

Who’s Set to Benefit? The US Stock Shortlist

Goldman Sachs and China Merchants Securities both highlight the leading names across the supply chain. According to the latest public disclosures and Nvidia’s own supplier documentation, major US-listed players include:

Semiconductor & Power Solution Providers:

Analog Devices, Inc.(ADI.US)

Alpha and Omega Semiconductor Limited(AOSL.US) 

Monolithic Power Systems, Inc.(MPWR.US) 

Navitas Semiconductor Corp Ordinary Shares - Class A(NVTS.US) 

ON Semiconductor Corporation(ON.US) 

Power Integrations, Inc.(POWI.US) 

STMicroelectronics NV Sponsored ADR RegS(STM.US) 

Texas Instruments Incorporated(TXN.US) 

INFINITY ENERGY RESOURCES INC(IFNY.US) 

RENESAS ELECTRONICS CORP(RNECY.US) 

ROHM CO LTD(ROHCY.US) 

Power System Components:

Flex Ltd(FLEX.US) GE Vernova Inc.(GEV.US)

Data Center Power System Suppliers:

Eaton Corp. Plc(ETN.US) GE Vernova Inc.(GEV.US) VERTIV HOLDINGS LLC(VRT.US)

Vertiv has just launched its space-saving, cost-efficient 800 VDC MGX reference architecture, a complete power and cooling infrastructure solution for next-gen centers.

NVIDIA Corporation(NVDA.US) projects that, over the long term, its 800VDC architecture could reduce total cost of ownership (TCO) by up to 30% through improvements in efficiency, reliability, and system design—along with lower equipment maintenance costs. Of course, this transition will require substantial upfront investment. The migration to 800VDC data centers is expected to coincide with the rollout of Nvidia’s Kyber rack architecture, with a target timeline set for 2027. Goldman Sachs forecasts that commercial applications of these technologies will begin to deliver significant economies of scale around 2028. For investors, keeping an eye on the leading companies shaping next-generation data center power standards could be one of the best ways to position for the 2026 AI infrastructure upgrade cycle.