STMicroelectronics expands 800 VDC AI datacenter power conversion portfolio with new 12V and 6V architectures in collaboration with NVIDIA

STMicroelectronics (NYSE: STM) expanded its 800 VDC datacenter power-conversion portfolio on March 17, 2026, adding 800 VDC to 12V and 800 VDC to 6V architectures developed to the NVIDIA 800 VDC reference design.

Rhea-AI Summary

STMicroelectronics (NYSE: STM) expanded its 800 VDC datacenter power-conversion portfolio on March 17, 2026, adding 800 VDC to 12V and 800 VDC to 6V architectures developed to the NVIDIA 800 VDC reference design. These complement ST's prior 800 VDC to 50V stage to support gigawatt-scale AI compute.

ST highlights system-level efficiency, reduced copper and losses, and prior prototype performance above 98% efficiency and > 2,600 W/in³ power density (Oct 2025).

Positive

• Adds 800 VDC to 12V and 6V architectures (Mar 17, 2026)
• Prototype 800 VDC GaN LLC achieved >98% efficiency (Oct 2025)
• Prototype power density exceeded 2,600 W/in³ at 50V (Oct 2025)

Negative

• None.

Argus Mar 17 session

+0.99% close to close Open Argus

Details

News Market Reaction – STM

On Mar 17, the day this news came out, STM closed 0.99% above the previous close.

Data tracked by StockTitan Argus for the Mar 17 session.

Key Figures

Input bus voltage: 800 VDC Converter output: 12V Converter output: 6V +4 more

Input bus voltage

800 VDC

AI datacenter power distribution architecture

Converter output

12V

New 800 VDC to 12V architecture for AI accelerators

Converter output

6V

New 800 VDC to 6V GPU-near power path

Intermediate bus

50V

Previously introduced 800 VDC to 50V stage

Switching frequency

1 MHz

GaN-based LLC converter prototype operating directly from 800 V

Efficiency

over 98%

GaN-based LLC converter from 800 V to 50V

Power density

2,600 W/in³

Smartphone-sized 50V converter prototype footprint

Historical Context

5 past events · Latest: Mar 10

5 events

1. Mar 10

   UWB SoC launch

   24h Move

   -0.0%

   Launch of ST64UWB ultra‑wideband SoC family for automotive and smart devices.

2. Mar 09

   AI photonics ramp

   24h Move

   +6.9%

   High-volume production of PIC100 silicon photonics for hyperscaler AI transceivers.

3. Mar 05

   New MCU family

   24h Move

   -0.1%

   Introduction of STM32C5 entry-level MCU series on 40 nm with Cortex-M33 cores.

4. Mar 02

   Wearables design win

   24h Move

   -5.6%

   ST sensors and secure wireless adopted in Qualcomm Snapdragon Wear Elite platform.

5. Mar 02

   AI investor calls

   24h Move

   -0.4%

   Announcement of March investor webcasts on cloud AI and intelligent sensing.

24h Move is the share-price change in the day after each event; other market factors may also have contributed.

Key Terms

sic, ir drop, gpu

3 terms

sic technical

"combine ST technologies across power semiconductors (silicon, SiC, GaN)"

Sic is a short notation placed immediately after a quoted word or passage to show the text is reproduced exactly as it appeared in the original source, even if it contains a typo, odd phrasing, or factual error. For investors, it signals that any mistake or unusual wording came from the original speaker or document, not from the publisher—like an editor saying “this is how it was said,” which helps with accuracy, responsibility, and legal clarity.

ir drop technical

"closer to the GPU, minimizing IR drop and improving response"

An IR drop is the loss of electrical voltage that happens when current flows through resistance in a circuit, described by Ohm’s law as current times resistance. For investors, it matters because excessive IR drop in power delivery systems, chips or batteries can reduce device performance, cause overheating, shorten lifespan, and force costly redesigns or recalls — similar to how a narrow pipe reduces water pressure at the tap.

gpu technical

"move the 6V bus closer to the GPU. This reduces copper usage"

A GPU (graphics processing unit) is a specialized computer chip designed to handle many calculations at once, originally for rendering images and video but now widely used for tasks like artificial intelligence, data analysis and high-performance computing. Investors watch GPU demand and prices because strong sales often signal growth for chip makers and their customers, affect profit margins and capital spending, and can forecast wider trends in gaming, AI adoption and cloud services.

AI-generated analysis. How Rhea-AI works. Not financial advice.

See more from StockTitan in Google Search and AI answers. Adds StockTitan as a preferred source · opens Google

Add on Google Not now

STMicroelectronics expands 800 VDC AI datacenter power conversion portfolio with new 12V and 6V architectures in collaboration with NVIDIA

• 12V, 6V solutions complement ST’s existing 800 VDC to 50V stage solution and are highlighted during NVIDIA GTC 2026
• ST now provides a complete portfolio for the 800 VDC power distribution inside gigawatt-scale compute infrastructure
• Solutions leverage ST power, analog and mixed-signal with custom design at both chip and package levels
  
  

Geneva, Switzerland — March 17, 2026 — STMicroelectronics (NYSE: STM), a global semiconductor leader serving customers across the spectrum of electronics applications, today announced the expansion of its 800 VDC power conversion portfolio with two new advanced architectures: 800 VDC to 12V and 800 VDC to 6V. Developed according to the NVIDIA 800 VDC reference design, these new power conversion stages complement the previously introduced 800 VDC to 50V solution. The rapidly emerging 800 VDC data center architecture enables higher energy efficiency, reduces power losses, and supports more scalable, high compute density, infrastructure for hyperscalers and AI compute.

“As AI infrastructure compute scale continues to expand fast, it requires higher voltage distribution and greater density, which can only be achieved with system-level innovation for each of the different AI server form factors,” said Marco Cassis, President, Analog, Power & Discrete, MEMS and Sensors Group Head of STMicroelectronics’ Strategy, System Research and Applications, Innovation Office at STMicroelectronics. “With these new converters for 800 VDC power distribution, ST brings a complete set of solutions to support the deployment of gigawatt-scale compute infrastructure with more efficient, scalable, and sustainable power architectures.”

A complete 800 VDC ecosystem for the different AI server form factors

The expansion to 12V and 6V output stages reflects the industry move toward different server architectures requiring different power delivery topologies depending on GPU generation, server height, form factor, and thermal envelope for large-scale training clusters, inference farms, and high-density AI infrastructures. The 50V, 12V, and 6V intermediate DC buses will all coexist in AI data centers depending on rack density, GPU configuration, and cooling strategy.
The new 800 VDC to 12V converter enables high-efficiency distribution from rack-level power shelves directly to the voltage domains that feed advanced AI accelerators.

The new 800 VDC to 6V path allows OEMs to reduce the number of conversion stages and move the 6V bus closer to the GPU. This reduces copper usage, minimizes resistive losses, and improves transient performance, a critical differentiator for large-scale training clusters.

Back in October 2025, STMicroelectronics introduced a fully integrated prototype power‑delivery system showcasing a compact GaN‑based LLC converter operating directly from 800 V at 1 MHz with over 98% efficiency and exceptional power density in a smartphone‑sized footprint exceeding 2,600 W/in³ at 50 V.

The three solutions combine ST technologies across power semiconductors (silicon, SiC, GaN), analog and mixed-signal, and microcontrollers.

Technical highlights of the new 12V and 6V architectures

Direct 800 VDC to 12V high-efficiency conversion:

• Eliminates the traditional 54V intermediate stage, reducing conversion steps and system-level losses.
• Enables higher rack-level efficiency, lower copper usage, and simplified integration for future GPU generations.
• Includes newly developed high-density power delivery board (PDB) achieving efficiency targets exceeding the sum of previous two-stage conversion paths.
  
  

800 VDC to 6V architecture for GPU-nearing conversion:

• Is designed for system builders who require power stages closer to the GPU, minimizing IR drop and improving response under fast load transients.
• Completes the topology portfolio for servers with ultra dense GPU configurations.

Additional technical information at https://blog.st.com/800-v-hvdc-data-center/

About STMicroelectronics

At ST, we are 48,000 creators and makers of semiconductor technologies mastering the semiconductor supply chain with state-of-the-art manufacturing facilities. An integrated device manufacturer, we work with more than 200,000 customers and thousands of partners to design and build products, solutions, and ecosystems that address their challenges and opportunities, and the need to support a more sustainable world. Our technologies enable smarter mobility, more efficient power and energy management, and the wide-scale deployment of cloud-connected autonomous things. We are on track to be carbon neutral in all direct and indirect emissions (scopes 1 and 2), product transportation, business travel, and employee commuting emissions (our scope 3 focus), and to achieve our 100% renewable electricity sourcing goal by the end of 2027. Further information can be found at www.st.com.

For further information, please contact:

INVESTOR RELATIONS
Jérôme Ramel
EVP Corporate Development & Integrated External Communication
Tel: +41 22 929 59 20
jerome.ramel@st.com

MEDIA RELATIONS
Alexis Breton
Group VP Corporate External Communications
Tel: +33 6 59 16 79 08
alexis.breton@st.com

Attachments

• T4766D -- Mar 17 2026 -- ST expands 800 VDC AI datacenter power conversion _FINAL FOR PUBLICATION
• ST expands 800 VDC AI datacenter power conversion portfolio

FAQ

AI-generated questions and answers. How Rhea-AI works. Not financial advice.

What did STMicroelectronics announce about 800 VDC power conversion (STM) on March 17, 2026?

ST announced new 800 VDC to 12V and 800 VDC to 6V architectures for AI datacenters. According to the company, these complement its existing 800 VDC to 50V stage and follow the NVIDIA 800 VDC reference design.

How do ST's new 12V and 6V architectures affect datacenter efficiency for STM shareholders?

They aim to reduce conversion stages and system losses, improving rack-level efficiency. According to the company, moving to 6V can reduce copper usage, minimize resistive losses, and improve transient response near GPUs.

What technical performance did ST report for its 800 VDC prototype (STM) in October 2025?

The prototype GaN-based LLC converter operated from 800 V at 1 MHz with over 98% efficiency. According to the company, it also demonstrated power density exceeding 2,600 W/in³ in a compact footprint.

Why is an 800 VDC to 6V path important for AI GPU architectures (STM)?

A 6V bus moves conversion closer to the GPU, reducing IR drop and improving transient response. According to the company, this minimizes resistive losses and supports ultra-dense GPU server configurations.

Will ST's 12V and 6V solutions replace the 50V intermediate bus in AI data centers (STM)?

Not necessarily; the company expects 50V, 12V, and 6V buses to coexist depending on rack density and GPU configurations. According to the company, choices depend on server form factor, thermal envelope, and cooling strategy.

Does ST provide a commercial availability date for the new 12V and 6V converters (STM)?

No commercial shipment date was provided in the announcement. According to the company, the release describes architecture expansion and technical highlights without specifying product launch timelines.