-
Innoscience has announced its collaboration with NVIDIA to support the implementation of the 800 VDC rack power architecture, a major breakthrough in power distribution for AI data centers. Similar to the transition from 400 V to 800 V systems in electric vehicles, this new architecture dramatically increases efficiency, power density, and sustainability by reducing current flow sixteenfold compared to traditional 48 V systems. The result is significantly lower resistive losses, reduced copper consumption, and lower overall CO₂ emissions.
Current 48 V data center systems are approaching physical and thermal limits, with high copper demand and nearly half of total power lost to heat dissipation. The 800 VDC system overcomes these constraints, enabling scalability from kilowatt to megawatt levels for future AI clusters with hundreds of GPUs. However, achieving such high power density and efficiency—especially converting from 800 V down to 1 V—requires advanced semiconductor technologies such as gallium nitride (GaN).
Innoscience’s third-generation GaN devices are designed to meet the demanding requirements of this new power paradigm. Operating at switching frequencies near 1 MHz, these devices significantly reduce the size of magnetic and capacitive components while improving overall system efficiency.
Key advantages of Innoscience’s third-generation GaN include:
- Up to 80% reduction in driver losses and 50% lower switching losses compared to SiC devices, yielding a 10% decrease in total system power loss.
- Twice the power density on the 54 V output side, requiring only 16 GaN devices instead of 32 Si MOSFETs for equivalent conduction performance.
- 70% reduction in switching losses and a 40% increase in power output compared to silicon-based architectures.
- Scalability to support next-generation GPU platforms with improved dynamic response and reduced capacitor costs.
As the world’s only integrated GaN manufacturer (IDM) producing devices from 1200 V to 15 V, Innoscience offers a complete GaN-based conversion chain from 800 V to 1 V, delivering an all-GaN power solution across every stage of the data center power architecture.
Innoscience’s GaN devices are also proven for long-term reliability, passing extended high-temperature and stress tests, including 2000-hour dynamic HTOL and 175°C endurance validation. These tests confirm datacenter-grade durability with expected lifetimes exceeding 20 years.
With its third-generation GaN technology and collaboration with NVIDIA, Innoscience is helping drive the shift from kilowatt to megawatt AI racks—ushering in a new era of ultra-efficient, high-performance, and sustainable AI computing infrastructure.
Original – Innoscience Technology
-
Navitas Semiconductor has announced new progress in developing high-performance gallium nitride (GaN) and silicon carbide (SiC) power devices designed to enable NVIDIA’s 800 VDC data center architecture for next-generation AI factory computing platforms.
With AI workloads driving data center power consumption into the megawatt range, traditional 54 V power distribution systems are reaching their limits. The shift to 800 VDC architecture offers higher efficiency, reduced copper use, simplified thermal management, and compatibility with global low-voltage DC standards. The new system allows direct conversion from 13.8 kVAC utility power to 800 VDC, eliminating multiple conversion stages and improving overall system reliability.
Navitas’ GaNFast™ and GeneSiC™ technologies power every stage of the AI data center — from grid connection to GPU-level conversion. The company’s new 100 V GaN FETs deliver high power density and thermal performance in dual-sided cooled packages, optimized for DC-DC conversion on GPU boards. Manufactured using a 200 mm GaN-on-Si process in partnership with Power Chip, these devices support scalable, high-volume production.
In addition, Navitas’ 650 V GaN portfolio introduces new high-power GaN FETs and GaNSafe™ power ICs, which integrate drive, sensing, protection, and control functions for enhanced safety and robustness. GaNSafe™ features ultra-fast short-circuit protection, 2 kV ESD tolerance, programmable slew-rate control, and a simple 4-pin configuration for easy implementation.
Navitas’ GeneSiC™ SiC technology, based on its proprietary trench-assisted planar structure, provides high-speed, cool-running performance and wide voltage coverage up to 6,500 V. These devices are already being used in grid-tied inverters, large-scale energy storage systems, and U.S. Department of Energy (DoE) projects.
“As NVIDIA drives transformation in AI infrastructure, we’re proud to support this shift with GaN and SiC power solutions that enable the efficiency, scalability, and reliability required by next-generation data centers,” said Chris Allexandre, President and CEO of Navitas. “The move from legacy 54 V systems to 800 VDC is not just an evolution — it’s a complete transformation of data center power.”
Navitas’ innovations in GaN and SiC technologies reflect its expanded focus beyond mobile and consumer markets toward powering megawatt-scale AI factories, industrial platforms, and smart energy infrastructure with high-efficiency, high-density semiconductor solutions.
Original – Navitas Semiconductor
-
ROHM has published a new white paper exploring advanced semiconductor solutions that enable the next generation of AI data centers powered by the 800 VDC architecture. The paper highlights ROHM’s role as a leading innovator in wide bandgap technologies and system-level power design for large-scale, energy-efficient computing infrastructure.
Developed as part of ROHM’s ongoing collaboration with NVIDIA and other industry partners, the white paper outlines strategies for implementing 800 VDC power distribution across AI data centers—a shift expected to transform data center design by enhancing efficiency, scalability, and sustainability.
The 800 VDC architecture supports the evolution of gigawatt-scale AI factories by significantly improving power density and reducing conversion losses. ROHM’s broad portfolio of silicon (Si), silicon carbide (SiC), and gallium nitride (GaN) power devices, combined with its analog IC expertise, positions the company to deliver optimized solutions for each stage of power conversion.
The paper details how the traditional AC-DC conversion process, typically performed inside server racks, can be relocated to a centralized power rack under the 800 VDC system. This approach simplifies design, improves thermal performance, and allows higher-density configurations for GPU-heavy AI workloads.
ROHM’s EcoSiC™ and EcoGaN™ device families are featured prominently as key enablers of this transition. The EcoSiC™ series provides industry-leading low on-resistance and top-side cooling modules for high-power AI servers, while the EcoGaN™ series integrates GaN performance with ROHM’s proprietary analog IC technologies, including Nano Pulse Control™, enabling fast, precise, and stable high-frequency operation.
Through collaborations with NVIDIA, Delta Electronics, and data center operators, ROHM continues to drive innovation in wide bandgap semiconductors and analog control technologies. Its new white paper underscores the company’s commitment to building efficient, reliable, and sustainable AI data center infrastructure based on 800 VDC power delivery systems.
Original – ROHM
-
Infineon Technologies has announced its support for NVIDIA’s 800 VDC power architecture, introduced at Computex 2025, marking a major step toward creating more efficient, scalable, and serviceable AI data centers. As the power demands of artificial intelligence infrastructure continue to grow exponentially, the move from traditional 54 V systems to centralized 800 VDC architectures is becoming essential to improve energy efficiency, reduce losses, and increase overall system reliability.
Adam White, Division President of Power & Sensor Systems at Infineon Technologies, emphasized the company’s role in shaping this transformation. “There is no AI without power,” he stated. “We are working with NVIDIA on intelligent power systems that not only meet the demands of future AI data centers but also minimize system downtimes. By driving the shift to high-density, reliable, and safe 800 VDC architectures, we are redefining how power is delivered to AI infrastructure and maximizing the value of every watt.”
Infineon’s collaboration with NVIDIA focuses on ensuring both safety and serviceability in 800 VDC-powered systems. One key innovation is the integration of hot-swap controller functionality, which enables server boards to be safely replaced or maintained while the rest of the rack continues operating. This solution, based on Infineon’s CoolSiC™ JFET technology, allows controlled pre-charging and discharging of server boards, preventing electrical hazards and eliminating costly downtime.
As AI data center power requirements rise — with rack power expected to reach 500 kilowatts and potentially 1 megawatt by the end of the decade — Infineon is developing next-generation power conversion systems that combine its Intermediate Bus Converter (IBC) and high-frequency gallium nitride (GaN) switching technologies. These solutions support efficient two- and three-stage power conversion from the grid to the server core, achieving up to 98 percent efficiency per conversion stage.
The company’s holistic approach extends beyond power conversion to include protection and control components that enhance reliability and sustainability. By leveraging silicon carbide (SiC), GaN, and silicon technologies, Infineon provides a complete portfolio of semiconductor solutions to enable safe and efficient megawatt-scale AI server racks.
Infineon’s experts will further discuss power conversion solutions for future server boards operating directly from high-voltage DC at the OCP Global Summit 2025 and explore advancements in AI data center power systems at OktoberTech Silicon Valley 2025.
With these developments, Infineon continues to position itself at the forefront of the semiconductor industry’s drive toward powering the AI revolution with smarter, more sustainable, and higher-performance energy systems.
Original – Infineon Technologies
-
Texas Instruments (TI) has introduced a new suite of power-management technologies and design resources to address the increasing energy demands of artificial intelligence (AI) data centers. The announcement was made ahead of the Open Compute Project (OCP) Global Summit in San Jose, California, where TI will showcase its latest innovations supporting the industry’s transition from 12 V and 48 V power systems to high-efficiency 800 VDC architectures.
As AI workloads grow and IT rack power approaches the megawatt scale, TI is collaborating with NVIDIA to develop advanced power-management devices optimized for the emerging 800 VDC data center standard. These new solutions aim to deliver higher efficiency, improved power density, and scalable energy conversion for next-generation AI computing infrastructure.
Among the highlights of TI’s announcement are several key resources and reference designs:
- White paper: “Power delivery trade-offs when preparing for the next wave of AI computing growth.” The paper explores design strategies for high-efficiency, high-density power delivery systems, emphasizing the technical benefits and challenges of adopting 800 VDC architectures.
- 30 kW AI server power-supply reference design: A dual-stage design featuring a three-phase, three-level flying capacitor PFC converter paired with dual delta-delta three-phase LLC converters. The system can be configured for a single 800 V output or split supplies to meet demanding AI workloads.
- Dual-phase smart power stage (CSD965203B): Delivers up to 100 A per phase and offers the highest peak power density in its class, allowing designers to scale power delivery efficiently across compact PCBs.
- Dual-phase smart power module (CSDM65295): Provides up to 180 A of peak output current in a 9 mm × 10 mm × 5 mm package, integrating two inductors and supporting trans-inductor voltage regulation for superior thermal and electrical performance.
- GaN intermediate bus converter (LMM104RM0): Offers up to 1.6 kW of output power with over 97.5 percent conversion efficiency in a compact quarter-brick form factor, supporting high light-load performance and active current sharing between modules.
According to Chris Suchoski, Sector General Manager for Data Centers at TI, the evolution of AI data centers requires scalable and efficient power infrastructures. “With the growth of AI, data centers are evolving from simple server rooms to highly sophisticated power infrastructure hubs,” he said. “Scalable power infrastructure and higher power efficiency are essential to meet these demands and drive future innovation. With TI devices, designers can build next-generation solutions that enable the transition to 800 VDC.”
TI’s participation at the OCP Global Summit will include live demonstrations, technical sessions, and presentations exploring advanced data center power architectures. The company continues to position itself as a key technology partner in developing efficient, reliable, and scalable power solutions that support the ongoing transformation of AI data centers worldwide.
Original – Texas Instruments
-
Renesas Electronics Corporation has announced its support for NVIDIA’s 800 VDC power architecture, introducing advanced GaN-based power technologies designed to enhance efficiency, scalability, and energy optimization in next-generation AI data centers.
As AI workloads accelerate and data center power consumption reaches hundreds of megawatts, the industry is moving toward higher-voltage, direct-current systems that reduce energy loss and simplify power distribution. Wide bandgap semiconductors such as gallium nitride (GaN) are emerging as the key enabler of this transformation, offering faster switching speeds, lower conduction losses, and superior thermal performance compared to traditional silicon devices.
Renesas’ GaN-based power solutions are engineered to support efficient DC/DC conversion across a wide range of operating voltages, from 48 V up to 400 V, with the capability to scale to 800 V in stacked configurations. Using the LLC Direct Current Transformer (LLC DCX) topology, these converters achieve efficiencies of up to 98 percent while maintaining high power density and compact form factors.
For the AC/DC front end, Renesas employs bi-directional GaN switches that simplify rectifier design and increase system-level efficiency. The company’s broader power component lineup—including REXFET MOSFETs, controllers, and drivers—complements these converters, providing a complete power delivery ecosystem for high-performance AI infrastructure.
“AI is transforming industries at an unprecedented pace, and the power infrastructure must evolve just as quickly to meet the explosive power demands,” said Zaher Baidas, Senior Vice President and General Manager of Power at Renesas. “Renesas is powering the future of AI with scalable, high-density energy solutions built on our advanced GaN and MOSFET technologies, ensuring the performance and efficiency required for the next wave of AI innovation.”
Renesas has also released a white paper detailing its approach to 800 VDC power distribution and conversion, exploring how its GaN-based solutions support the evolution of energy-efficient, large-scale AI data center infrastructure.
Original – Renesas Electronics
-
STMicroelectronics has revealed a complete prototype of its new power delivery system designed to support NVIDIA’s 800 VDC architecture for next-generation AI data centers. The company’s announcement underscores its leadership in developing semiconductor technologies that meet the rising power and efficiency demands of large-scale AI computing infrastructure.
As AI workloads grow rapidly, traditional 54 V power distribution systems are reaching their limits. The shift to 800 VDC architectures enables megawatt-scale compute racks that are more efficient, require less copper, and simplify overall system design. STMicroelectronics is contributing to this transition with a portfolio that integrates silicon carbide (SiC), gallium nitride (GaN), and silicon technologies optimized for high-voltage, high-efficiency applications.
At the OCP Global Summit 2025, ST presented a major development milestone: a compact 12 kW GaN-based LLC power delivery board roughly the size of a smartphone. Operating from an 800 V input and switching at 1 MHz, the prototype achieved more than 98 percent efficiency and a record power density of over 2,600 W/in³ at 50 V output.
The new system addresses key design challenges in power density, thermal management, efficiency, and reliability—critical factors for deploying megawatt-scale AI compute systems while lowering infrastructure complexity and cost.
STMicroelectronics’ achievement represents a significant step forward in enabling high-performance, energy-efficient power delivery solutions for the emerging generation of hyperscale AI data centers built on 800 VDC architectures.
Original – STMicroelectronics
-
EPC has introduced a new high-efficiency power converter designed to accelerate the adoption of 800 VDC power distribution in next-generation AI data centers.
As artificial intelligence workloads continue to expand, future data centers—often referred to as AI factories—will require megawatt-scale rack-level power delivery systems. To meet these demands, EPC has developed a compact, low-cost, 6 kW GaN-based converter that steps down 800 VDC to 12.5 VDC using an Input Series Output Parallel (ISOP) topology.
The converter delivers multiple key advantages:
- High power density: The design occupies less than 5,000 mm² of board space and measures only 8 mm in height, making it ideal for space-constrained AI server boards.
- High efficiency: By converting power close to the load, the system minimizes bussing losses and enhances overall energy efficiency.
- Simplified architecture: Moving directly from AC to 800 VDC at the rack level and then stepping down to 12.5 VDC at the board eliminates unnecessary conversion stages, improving scalability and reducing system complexity.
EPC’s GaN-based approach supports the evolution of AI data centers toward more compact, efficient, and sustainable infrastructure.
“GaN is an essential technology for the 800 VDC ecosystem,” said Alex Lidow, CEO of EPC. “Our collaboration with NVIDIA focuses on delivering compact, efficient, and cost-effective board-level power conversion to support the next generation of AI factories operating at gigawatt scale.”
The development underscores EPC’s commitment to advancing GaN technology as a cornerstone for future high-voltage, high-efficiency power architectures in AI and high-performance computing environments.
Original – Efficient Power Conversion
-
Alpha and Omega Semiconductor (AOS) has announced its full support for NVIDIA’s newly introduced 800 VDC power architecture, designed to power the next generation of AI data centers with megawatt-scale racks. The company’s portfolio of silicon carbide (SiC) and gallium nitride (GaN) power devices is strategically developed to meet the performance, efficiency, and density requirements of this next-generation infrastructure.
The transition from traditional 54 V to 800 VDC power distribution marks a major redesign in data center architecture. This shift reduces the number of power conversion stages, improves energy efficiency, lowers copper usage, and enhances system reliability. The new architecture also demands advanced wide bandgap semiconductor technologies capable of operating at higher voltages and switching frequencies while maintaining maximum efficiency.
Ralph Monteiro, Senior Vice President of Power IC and Discrete Product Lines at AOS, stated that the company is collaborating with NVIDIA to develop 800 VDC power semiconductors tailored for high-efficiency power delivery across all conversion stages—from AC-to-DC input to DC-to-DC rack-level conversion.
AOS’ wide bandgap expertise positions the company as a key technology enabler for this transition. Its SiC and GaN devices are optimized for each stage of the new 800 VDC architecture:
- High-voltage conversion: AOS’s SiC devices, such as the Gen3 AOM020V120X3 and the topside-cooled AOGT020V120X2Q, support direct conversion from 13.8 kV AC grid input to 800 VDC, simplifying the power chain and improving system efficiency.
- High-density DC-DC conversion: Inside the racks, AOS’s 650 V and 100 V GaN FETs, including the AOGT035V65GA1 and AOFG018V10GA1, enable compact, high-frequency converters that free up rack space and improve cooling performance.
- Packaging innovations: The company’s stacked-die MOSFETs and GaN devices, such as the AOPL68801, share common footprints, allowing designers flexibility in cost and efficiency trade-offs in secondary-side power conversion.
- Multiphase controllers: AOS also provides multi-rail 16-phase controllers for efficient voltage regulation from 54 V to 12 V and beyond, supporting AI processor power delivery.
By supplying key technologies for each conversion stage, AOS enables up to a 5 percent increase in overall power efficiency, a 45 percent reduction in copper usage, and significant improvements in thermal and maintenance performance.
AOS’s continued investment in SiC and GaN development underlines its commitment to driving the scalability, sustainability, and performance of next-generation AI data centers built on 800 VDC power infrastructure.
Original – Alpha and Omega Semiconductor
-
Power Integrations has unveiled new details on the capabilities of its PowiGaN gallium-nitride technology designed to power the next generation of AI data centers. The company outlined the performance and system-level advantages of its 1250 V and 1700 V PowiGaN devices in a white paper released at the 2025 OCP Global Summit in San Jose, where NVIDIA provided an update on the emerging 800 VDC data center architecture.
The paper demonstrates how Power Integrations’ high-voltage PowiGaN HEMTs enable compact, high-efficiency 800 VDC power conversion systems with efficiency exceeding 98 percent. The company’s 1250 V PowiGaN switch delivers superior performance compared to stacked 650 V GaN FETs and conventional 1200 V silicon carbide (SiC) devices, offering greater power density, lower switching losses, and proven field reliability.
Power Integrations also highlighted its InnoMux-2 EP integrated circuits as an advanced solution for auxiliary power systems in 800 VDC data center designs. Featuring an integrated 1700 V PowiGaN switch, the device supports input voltages up to 1000 VDC and achieves over 90 percent efficiency in compact, fanless liquid-cooled architectures.
According to Roland Saint-Pierre, vice president of product development at Power Integrations, the industry’s move toward 800 VDC architectures addresses rising power demands in AI data centers by simplifying rack designs, improving space utilization, and reducing copper use. He emphasized that the company’s 1250 V and 1700 V PowiGaN devices are ideally suited to meet the efficiency, reliability, and power-density requirements of these high-performance systems.
Power Integrations is currently the only company supplying 1250 V and 1700 V GaN switches in high-volume production. Since introducing its first GaN ICs in 2018, the company has shipped more than 175 million GaN devices across a wide range of applications, including data centers, electric vehicles, and fast chargers.
The introduction of PowiGaN technology for 800 VDC data centers marks a significant step toward more efficient, scalable, and sustainable power conversion systems to support the rapidly growing energy demands of AI computing infrastructure.
Original – Power Integrations
