Navitas Targets AI Data Centers With High-Efficiency GaN Power Tech
The semiconductor firm is deploying GaN and SiC technologies to reduce energy consumption and heat in AI GPU clusters.
Navitas Semiconductor is aggressively positioning its Gallium Nitride (GaN) and Silicon Carbide (SiC) power technologies to capture the rapidly expanding AI data center market. The move addresses critical power and thermal challenges created by the surge in AI GPU and accelerator deployments, where traditional silicon-based components are reaching their physical limits.
To meet the demands of AI infrastructure, Navitas provides GaN and SiC power ICs and MOSFETs designed to increase power density and efficiency within server power supply units (PSUs). The company has developed PSU reference designs in 3.2kW, 4.5kW, and 8.5kW capacities. According to company data, these designs achieve 80 PLUS 'Ruby' certification, delivering efficiencies between 96.5% and 98%.
The Shift to Wide Bandgap Semiconductors
This strategic push comes as AI workloads drive a massive increase in data center power demand. Traditional silicon struggles to handle the heat and energy requirements of modern AI clusters, creating a market opening for Wide Bandgap (WBG) semiconductors. Unlike standard silicon, GaN and SiC offer superior thermal management and efficiency. This allows data centers to pack more computing power into the same physical footprint while reducing wasted energy and operational costs.
Industrial Pivot and Market Impact
For Navitas, this represents a fundamental shift in its business model. The company is strategically pivoting away from a revenue mix heavily dependent on consumer electronics, such as mobile chargers, toward high-growth industrial and enterprise sectors. By integrating into the AI infrastructure and electric vehicle (EV) supply chains, Navitas is moving into higher-margin, long-term enterprise contracts.
If the company successfully penetrates the AI data center power supply chain, the transition could significantly alter its long-term valuation. Moving from a consumer-centric portfolio to an industrial-grade infrastructure provider typically commands a different market multiple and provides more stable, scalable growth trajectories compared to the volatile consumer electronics market.
Future Outlook
Industry observers are now watching how quickly these reference designs transition into mass-market adoption. While the technical efficiency of GaN is proven, the scale of deployment across major cloud providers remains the key metric for success. Further developments in 800V architectures and deeper partnerships with GPU manufacturers will likely determine if Navitas can establish a dominant footprint in the AI power ecosystem as the industry scales toward next-generation power requirements.