The AI boom is creating an unprecedented electricity crisis, but the real bottleneck isn't computing power anymore,it's the aging power grid that delivers it. While investors focus on GPU shortages and chip performance, a quieter infrastructure race is unfolding behind the scenes. Companies like Navitas Semiconductor are betting that the utilities and grid operators upgrading their systems to handle AI's massive power demands represent an equally lucrative opportunity as the data centers themselves.

Why Is Grid Infrastructure Suddenly Critical for AI?

Traditional electrical grids were designed decades ago for a different era of computing. Today's AI data centers operate at scales that strain these aging systems. As hyperscale facilities proliferate across the United States, the demand for electricity is growing so rapidly that utilities cannot keep pace with traditional infrastructure upgrades alone.

The challenge extends beyond simply adding more power lines. Modern AI data centers require not just more electricity, but more efficient ways to convert, distribute, and manage that power at unprecedented scales. This is where advanced semiconductor technologies come into play. Navitas Semiconductor, which has attracted attention for its partnerships with NVIDIA and its gallium nitride (GaN) technology for data center power systems, is now positioning itself as a critical player in grid modernization as well.

On its latest earnings call, Navitas management revealed that its AI infrastructure segment, which includes both data centers and grid infrastructure, grew 50 percent sequentially. Chief Executive Officer Chris Allexandre emphasized that these two markets are not separate opportunities,they are fundamentally linked. As AI data centers proliferate and electricity consumption rises, utilities need more efficient solutions capable of handling higher power loads.

What Technologies Are Solving the Grid Problem?

Navitas is leveraging silicon carbide (SiC) technology to address grid modernization challenges. The company's 2.3 kilovolt and 3.3 kilovolt SiC modules are designed for applications including battery energy storage systems, utility solar farms, and solid-state transformers. These advanced power-conversion technologies represent a significant upgrade from traditional transformers, which were designed for a different era and are increasingly stretched by today's power requirements.

The company's 250 kilowatt solid-state transformer demonstration uses GeneSiC technology for scalable 800-volt direct current distribution. This represents a fundamental shift in how power is managed at the grid level. Solid-state transformers offer greater efficiency and flexibility compared to traditional electromagnetic transformers, making them ideal for integrating renewable energy sources and managing the variable power demands of AI infrastructure.

How Are Companies Positioning Themselves in This Market?

• Navitas Semiconductor: Targeting both data center power systems through GaN technology and grid infrastructure through high-voltage SiC, positioning itself at the intersection of AI and grid modernization with a total serviceable addressable market of $3.5 billion by 2030.
• onsemi: Building a strong position in energy storage systems with market share approaching 60 percent in string energy storage systems and microgrid applications, expecting revenues from these segments to grow more than 40 percent year over year in 2026.
• STMicroelectronics: Making grid and energy infrastructure a strategic priority by investing heavily in silicon carbide technologies while expanding relationships with solar, battery storage, and power-conversion customers.

Navitas maintains a strong balance sheet with no debt and $221 million in cash and cash equivalents as of the first quarter of 2026, providing flexibility to continue investing in research and development without immediate financing pressure.

What Is the Market Opportunity for Grid Infrastructure?

Navitas sees the energy and grid segment alone representing a $1 to $1.8 billion serviceable market by 2030. The company puts its total serviceable addressable market at $3.5 billion by 2030, split roughly 50-50 between GaN and high-voltage SiC technologies, with a combined compound annual growth rate exceeding 60 percent.

Customer engagement in U.S. grid infrastructure accelerated in the first quarter of 2026. Navitas is already seeing interest from customers involved in grid-scale solar, megawatt power conversion, and other energy infrastructure projects, with adoption expected to build through 2026 and 2027 before accelerating further later in the decade.

The broader U.S. AI data center market itself is projected to reach $610.12 billion by 2032, up from $142.50 billion in 2026, representing a compound annual growth rate of 27.4 percent. Within this market, cooling solutions are projected to grow at the highest rate of 28.5 percent due to rising heat density from high-performance AI workloads, while hyperscale data centers are expected to hold the largest market share of 68.4 percent in 2032.

Why Are Investors Missing This Infrastructure Story?

Most investors focus on the AI narrative,the chips, the software, the applications. But the physical infrastructure required to support AI at scale represents an equally important investment thesis. Shares of Navitas have rallied 232 percent year to date compared with the industry's growth of 66 percent, yet the company trades at a forward price-to-sales ratio of roughly 97 times, significantly higher than the industry's 10 times.

"The AI and grid infrastructure are not separate investment themes. They are, in fact, two sides of the same story. AI is driving an unprecedented increase in power demand, while grid modernization is the response required to support that demand," according to analysis of Navitas' market positioning.

Navitas Semiconductor Market Analysis, Source 1

Because Navitas has exposure to both data center power systems through GaN and grid infrastructure through high-voltage SiC, it sits at the intersection of these trends. Investors focused solely on the AI narrative may be missing a second growth engine that could become increasingly important over the next several years.

The convergence of AI infrastructure and grid modernization represents one of the most significant infrastructure buildouts since the early internet boom. As data centers consume ever-larger amounts of electricity, the utilities and infrastructure providers that can efficiently manage and distribute that power will become as critical to the AI ecosystem as the semiconductor companies building the chips themselves.