Navitas (NVTS) soars on Nvidia power partnership as AI data centers chase the next bottleneck

Navitas Semiconductor Corporation (NASDAQ: NVTS), a Torrance, California maker of advanced power semiconductors, surged more than 20 percent after revealing a collaboration with Nvidia’s MGX ecosystem to develop 800-volt direct current power architectures for artificial intelligence data centers. The stock jumped from a prior close of 25.86 dollars to trade in the low-to-mid 30s, adding roughly 1.3 billion dollars in market value in a single morning, as investors embraced the idea that power delivery is becoming the next critical constraint in the AI buildout. Navitas, which specializes in gallium nitride and silicon carbide power chips, is showcasing an 800-volt-to-6-volt power delivery board at the Computex conference in Taipei as part of Nvidia’s AI Factory MGX platform. The move extends an extraordinary run for one of the market’s hottest AI-power names, which gained 61 percent in May alone. Yet the surge also sharpens an uncomfortable question, because Navitas reported just 8.6 million dollars of quarterly revenue, leaving its valuation resting almost entirely on a future that has not yet arrived.

What did Navitas announce with Nvidia and why did the stock surge over 20%?

The catalyst was inclusion in Nvidia’s flagship AI infrastructure ecosystem. Navitas announced it is collaborating within Nvidia’s MGX framework to support emerging 800-volt direct current rack architectures, and it participated in Nvidia’s partner ceremony in Taipei in late May. The association with Nvidia, the company at the center of the AI boom, instantly elevated Navitas in investors’ eyes.

The specific technology on display gave the news substance. At Computex, Navitas is showcasing an 800-volt-to-6-volt power delivery board built on its GaNFast gallium nitride technology, designed to deliver electricity to AI compute racks at high efficiency. The board reportedly achieves around 97.5 percent efficiency and very high power density, and it eliminates the traditional intermediate voltage conversion stage, simplifying the power architecture of AI servers.

The market reaction reflected both momentum and narrative. Trading volume ran well above average as the stock added roughly 1.3 billion dollars in market value, and retail sentiment shifted notably more positive. Investors increasingly view power management as the next major growth area in AI infrastructure, following the waves that lifted memory and networking names, and Navitas became the immediate vehicle for that thesis.

Why is power delivery becoming the next critical bottleneck in AI data centers?

The core idea driving the rally is that AI data centers are running into power constraints. Modern AI server racks consume enormous and rapidly growing amounts of electricity, and delivering that power efficiently from the grid to the chips has become a defining engineering challenge. As compute density rises, the power architecture that feeds it must evolve to keep pace.

The industry’s answer is higher voltage. The shift toward 800-volt direct current rack architectures, championed by Nvidia, is designed to handle the megawatt-scale power demands of next-generation AI factories more efficiently than older, lower-voltage designs. Higher voltage reduces energy losses and the physical bulk of power distribution, which is critical when every rack draws as much electricity as a small neighborhood.

This creates a genuine opportunity for power semiconductor specialists. Just as memory and networking emerged as bottlenecks that lifted companies addressing them, power conversion is now drawing investor attention as the next constraint, and the firms whose chips enable efficient high-voltage power delivery stand to benefit. Navitas has positioned itself squarely in that niche, which is why a showcase within Nvidia’s ecosystem carried such weight.

How do Navitas’s GaN and SiC chips fit the 800 VDC AI rack transition?

Navitas specializes in wide-bandgap semiconductors, the materials best suited to the new power demands. Its GaNFast gallium nitride and GeneSiC silicon carbide technologies handle higher voltages, switch faster, and run more efficiently than traditional silicon power chips, making them well matched to the requirements of high-voltage AI infrastructure.

The chips address the full power path. Navitas’s silicon carbide solutions support delivery of electricity from the grid into the data center through applications such as solid-state transformers and high-power three-phase power supply units, while its gallium nitride technology handles the final conversion down to the low voltages the chips consume. Spanning both ends of the power chain broadens the company’s potential role in an AI rack.

The value proposition is efficiency, density, and thermal performance. By enabling megawatt-scale racks with higher power density, smaller footprints, and better heat management, Navitas argues its technology helps accelerate the transition to more efficient AI infrastructure. If the 800-volt architecture becomes the industry standard, demand for the wide-bandgap chips that make it work would grow alongside the relentless expansion of AI data centers.

Why do skeptics question the Navitas valuation against its tiny revenue base?

The central concern is the gap between the stock and the financials. Navitas reported quarterly revenue of just 8.6 million dollars, a tiny figure against a market capitalization that the surge pushed into the billions, implying a price-to-sales multiple in the triple digits. A valuation of that magnitude prices in years of explosive growth that the company has yet to demonstrate.

The nature of the catalyst also warrants caution. Being featured in Nvidia’s MGX ecosystem showcase and collaborating on an architecture is meaningful validation, but it is not the same as a binding, large-scale supply contract with guaranteed revenue. The market is rewarding a narrative and a positioning win, and the leap from showcase to material, recurring sales is not assured.

Dilution is an additional overhang. Navitas filed a shelf registration in May allowing it to issue up to 250 million dollars in securities, giving it flexibility to raise capital, which is sensible for a small, cash-consuming company but dilutive to existing shareholders. Combined with the stock’s history of sharp swings on Nvidia-related headlines, this makes Navitas a high-risk, sentiment-driven name rather than one anchored by current fundamentals.

How is Navitas stock positioned after a parabolic multi-month run?

The recent move is part of a much larger surge. Navitas climbed 61 percent in May before the latest jump, and the stock has repeatedly posted large gains on news tied to its Nvidia relationship and to bullish data points on AI infrastructure spending. The shares have become one of the most volatile and closely watched AI-power plays in the market.

The momentum cuts both ways. A stock that rises this fast on narrative and partnership news can continue to climb as enthusiasm builds, but it is equally prone to sharp reversals when sentiment cools or when the gap between expectations and results becomes too wide. The elevated trading volume signals intense speculative interest rather than steady accumulation.

For investors, the setup is a high-stakes bet on a theme. The bull case is that power delivery is a real, growing bottleneck and that Navitas is positioned to capture a share of an enormous opportunity. The bear case is that the valuation has detached from a minuscule revenue base on the strength of a showcase rather than secured revenue, leaving the stock vulnerable. The truth will depend on whether the collaboration translates into the design wins and contracts that justify the price.

What competition, dilution and execution risks should investors weigh?

The first risk is competition. The power semiconductor market includes large, well-established players with deep customer relationships and substantial resources, and Navitas must win against them to capture the AI power opportunity. Being in Nvidia’s ecosystem helps, but it does not guarantee that Navitas, rather than a larger rival, wins the volume business.

The second risk is execution and scale. Navitas is a small company with limited revenue that must scale manufacturing, secure design wins, and convert technological promise into reliable, high-volume production for demanding data center customers. The path from a showcase board to widespread adoption involves qualification, reliability, and supply commitments that take time and carry risk.

The third risk is valuation and financing. With the stock priced for a future far larger than its current business, any disappointment in design wins, revenue ramp, or AI infrastructure spending could trigger a steep pullback, and the company may tap its shelf registration to raise dilutive capital. None of this is investment advice, and the power-delivery thesis driving the rally reflects a real engineering challenge in AI data centers. But Navitas now embodies the speculative edge of the AI infrastructure trade, where a partnership headline can add over a billion dollars in value to a company whose revenue remains a tiny fraction of its market capitalization, and the stock’s future hinges on closing that gap.

Key takeaways on what the Nvidia collaboration means for Navitas

• Navitas surged more than 20 percent after announcing a collaboration with Nvidia’s MGX ecosystem to develop 800-volt direct current AI data center power architectures.
• The rally added roughly 1.3 billion dollars in market value as investors embraced power delivery as the next AI infrastructure bottleneck.
• Navitas is showcasing a high-efficiency 800-volt-to-6-volt power delivery board at Computex as part of Nvidia’s AI Factory MGX platform.
• AI server racks consume rapidly growing power, driving a shift to 800-volt architectures championed by Nvidia for better efficiency and density.
• Navitas’s gallium nitride and silicon carbide chips handle higher voltages more efficiently than traditional silicon, fitting the new power path.
• The company reported just 8.6 million dollars in quarterly revenue, leaving a triple-digit price-to-sales multiple after the surge.
• The catalyst is a showcase and collaboration, not a binding large-scale supply contract, so anticipated revenue is not yet secured.
• A 250 million dollar shelf registration creates potential dilution for a small, cash-consuming company.
• The stock gained 61 percent in May and trades on momentum and narrative, making it highly volatile.
• The thesis depends on converting the Nvidia relationship into design wins and contracts that justify a valuation far above current sales.