Navitas Pivots to AI Power Infrastructure as Data Center Energy Demands Surge
The semiconductor firm is shifting from mobile charging to high-power AI delivery, targeting a $3.5 billion market by 2030.
The explosive growth of artificial intelligence has created a critical bottleneck in power delivery, sparking a strategic shift for semiconductor players. While Nvidia provides the compute power driving the boom, Navitas Semiconductor is positioning itself to solve the energy efficiency crisis inherent in hyperscale environments.
Navitas is currently executing a "Navitas 2.0" pivot, moving its revenue mix away from mobile charging and toward high-power AI data center infrastructure. The company reported Q2 2026 revenues of $10.5 million, marking a 22% sequential increase. More telling is the growth in its high-power markets, which have surged over 50% year-over-year. This trajectory is supported by a $450 million backlog of design wins, providing significant visibility into the company's expansion. To further cement its role in the ecosystem, Navitas showcased an 800V-to-6V GaNFast power delivery board specifically designed for the NVIDIA MGX platform at GTC 2026.
The Power Bottleneck
The AI surge has pushed traditional silicon-based power chips to their limits. The extreme energy requirements of AI training and inference have made legacy components inefficient, creating a demand for Wide Bandgap (WBG) semiconductors. Navitas utilizes Gallium Nitride (GaN) and Silicon Carbide (SiC) technologies, which allow for higher power density and superior efficiency compared to standard silicon. According to Navitas management, AI infrastructure growth is already occurring ahead of the full 800V transition, as current data center power requirements continue to climb.
Strategic Market Implications
This shift represents a diversification of the AI investment landscape, moving from the high-valuation compute layer to the essential infrastructure layer. By providing the "circulatory system" for AI's "brains," Navitas is targeting a serviceable addressable market (SAM) estimated at $3.5 billion by 2030 across high-power segments, including grid infrastructure and AI data centers. For hyperscalers, the adoption of these high-efficiency solutions is a primary lever for lowering the Total Cost of Ownership (TCO) of massive AI clusters.
Future Outlook
Navitas expects AI infrastructure to represent more than one-third of its total sales by the end of 2026. The company's ability to scale these high-power solutions will determine if it can capture a dominant share of the energy-efficiency spend. Investors and industry observers will be watching whether the current backlog of design wins converts into sustained revenue growth as the industry moves toward more aggressive power delivery standards.