Vertiv and ON Semiconductor both serve the AI data center market as Nvidia partners, but they target different phases of AI spending. Vertiv dominates the current build-out phase with power and cooling solutions; ON Semiconductor, cheaper on valuation, is betting on physical AI—inference at the edge in robots and electric vehicles—where it plans to offer an integrated stack of sensors, power management, and computing after acquiring Synaptics for $7 billion（約1.1兆円）. Industry observers expect edge AI to grow long-term after the initial massive cloud investment phase.
Summaries like this, in your inbox every morning.
Sign up free →What happened
Two AI infrastructure stocks—Vertiv, which supplies power and cooling for data centers, and ON Semiconductor, which makes power and sensing chips—trade at different valuations despite both being Nvidia partners. ON Semiconductor is pursuing a $7 billion（約1.1兆円） acquisition of edge solutions company Synaptics to build out its physical AI capabilities.
Why it matters
AI spending will likely shift from cloud-based model training and inference (phase one) to edge AI—inference performed by robots and electric vehicles near where data is collected (phase two). Vertiv leads in phase one, but ON Semiconductor is positioning itself for the long-term phase-two boom. ON Semiconductor's data center business is expected to double in 2026, and the Synaptics deal gives it sensors, power management, control technology, and edge computing in one package—a full stack for industrial robotics and autonomous systems.
What to watch
ON Semiconductor's data center business is forecast at $500 million（約800億円） of its nearly $6.5 billion（約1兆円） total revenue in 2026. The critical question is whether management can successfully integrate Synaptics and capitalize on the shift to physical AI infrastructure.
Vertiv and ON Semiconductor occupy different positions in the artificial intelligence infrastructure supply chain, despite both partnering with Nvidia on data center technology. Vertiv manufactures power, cooling solutions, and data center architecture—the backbone of today's hyperscaler build-out. ON Semiconductor makes power and sensing chips primarily for electric vehicles and industrial applications such as EV charging, renewable energy, and industrial automation. Its data center business, while growing, contributes only $500 million（約800億円） of its nearly $6.5 billion（約1兆円） projected 2026 revenue.
The article identifies two overlapping phases in AI infrastructure spending. The first phase is the AI build-out: hyperscalers train large language models and run inference—the computational step where AI models generate answers—in centralized cloud data centers. This phase is where Vertiv has clear, immediate exposure and where hyperscalers are currently deploying capital. The second phase is physical AI, where inference migrates to the edge. Rather than sending data to the cloud for processing, AI-powered robots or electric vehicles perform inference close to where data is collected. The article suggests that after the massive initial investment in phase one, spending in phase two will grow over the long term.
ON Semiconductor has begun repositioning for this second phase. Its data center business is set to double in 2026, and the company announced a $7 billion（約1.1兆円） acquisition of Synaptics, an edge solutions company, to close a gap in its physical AI portfolio. With this deal, ON Semiconductor will offer an integrated platform for industrial robots and similar devices: sensors to collect data, power management systems to efficiently drive the robot, control technology to command motors and motion, and—via Synaptics—edge AI computing to make real-time decisions. This vertical integration contrasts with Vertiv's narrower exposure to cloud infrastructure and indirect exposure to physical AI growth.
Valuation reflects the market's confidence in each narrative. Investors currently pay significantly more for Vertiv's near-term, high-certainty AI build-out exposure than for ON Semiconductor's longer-term physical AI bet. However, the article argues that ON Semiconductor is the more compelling option if one believes physical AI—autonomous driving, robotics, and industrial automation—is the future. ON Semiconductor trades cheaper and offers full-stack capability in the edge AI space, contingent on successful Synaptics integration.
The article frames a divergence in how two Nvidia partners will capture value as artificial intelligence infrastructure evolves. Vertiv benefits immediately from hyperscalers' mammoth current spending on data center build-out—power systems, cooling, and architecture. ON Semiconductor, by contrast, has historically served power and sensing chip markets in electric vehicles and industrial automation; its data center footprint today is small. Yet the article posits that AI spending patterns will shift fundamentally over time. Once the initial phase of cloud-based model training and inference reaches saturation, the growth engine moves to edge AI—inference running locally on autonomous robots, electric vehicles, and industrial equipment. This shift favors ON Semiconductor's existing expertise in power management and sensing, combined with the computational edge it will gain from Synaptics. The valuation gap reflects this temporal mismatch: investors pay a premium for Vertiv's near-term, certain exposure, while ON Semiconductor trades cheaper despite offering what the article characterizes as long-term upside. The outcome depends on the credibility of the phase-two thesis and ON Semiconductor's ability to integrate Synaptics and execute.
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytime
No comments yet. Be the first to share your thoughts!
Log in to join the discussion
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
