About TrendForce News

TrendForce News operates independently from our research team, curating key semiconductor and tech updates to support timely, informed decisions.

[News] Innoscience Advances AI Infrastructure Power as Part of the NVIDIA MGX Ecosystem



China’s gallium nitride (GaN) power semiconductors giant Innoscience, during the Open Compute Project (OCP) Global Summit in late October, showcased its GaN power solutions for next-generation AI infrastructure as a participant in the NVIDIA MGX ecosystem.

As AI compute scales, power has become a defining constraint on data-center performance. Every watt lost in conversion limits how much compute a rack can deliver. Innoscience’s GaN solutions address this bottleneck directly, enabling higher power density, lower energy losses and greater system efficiency across AI servers, data-center power supplies and emerging 800 VDC architectures.

Scale Built for AI Demand

According to Innoscience, the company operates the world’s largest vertically integrated GaN manufacturing platform and leads the industry in high-volume 8-inch GaN-on-silicon production. This combination of design, process and manufacturing under one roof gives hyperscalers and AI infrastructure providers what they need most as deployments accelerate: proven volume, supply-chain resilience and cost competitiveness at scale.

Participation in the NVIDIA MGX ecosystem reflects this readiness and positions Innoscience to support partners building the modular, high-performance platforms that underpin AI infrastructure, the company said.

A Fast-Growing Market

According to Yole Group’s Power GaN 2026 report, the global power GaN device market is projected to reach approximately US$3.5 billion by 2031, a 35% CAGR. Infrastructure and telecom applications, including AI data centers, are expected to grow faster still, at a 45% CAGR.

The broader GaN semiconductor devices market, spanning power electronics, RF, telecommunications infrastructure, automotive and industrial applications, is projected by Grand View Research to reach approximately US$28.9 billion by 2033.

With its manufacturing scale, broad product portfolio and established high-volume production, Innoscience is well positioned to support customers across this expanding landscape, it said.

(Photo credit: Innoscience)

Please note that this article cites information from NVIDIA MGX and Innoscience.



Get in touch with us