ON Semiconductor (ON.US), a major chip manufacturer focused on the automotive and industrial sectors, released its quarterly earnings and future outlook after the market close on Monday (early Tuesday morning Beijing time). The company's third-quarter revenue forecast range exceeded the consensus estimate of Wall Street analysts, highlighting the explosive growth of its AI data center "power engine" business. Surge in demand for power management chips used in AI data centers drove the stock price up more than 7% in after-hours trading.

Amid the AI infrastructure boom, the company's second-quarter results for the period ending July 3 were also strong. The Power Solutions Group (PSG) business, which includes automotive, industrial, and AI data center operations, posted the strongest growth. Since the beginning of the year, ON Semiconductor's stock has risen significantly, driven by the unprecedented wave of AI data center construction and the expected recovery in demand for analog chips and data center power chain components. Although the stock has corrected since July due to deleveraging in the global AI computing theme and the unwinding of extremely crowded positions, it is still up 50% since 2026. In contrast, the chip giant's stock fell 15% for the full year of 2025.

The surge in chip demand from AI is spreading from the "core computing chip (GPU/ASIC/HBM)" to the broader "data center power and signal chain," and the intensity of this spillover is accelerating significantly. This further validates the logic of the "AI end is power" bull market curve for the data center power chain. The seemingly "endless" demand for chips from AI training and inference is smoothly transitioning from AI chips and memory chips to the analog and power semiconductor segments, strongly driving the performance recovery of leaders like Texas Instruments, Infineon, and ON Semiconductor. The market interprets these strong results as evidence that the "analog chain is beginning to reap the super dividends of AI infrastructure."

The $7 Billion Acquisition and AI Data Center Ignite Power Chip Demand, Launching a New Growth Cycle for ON Semiconductor

Alongside its strong results and outlook, ON Semiconductor is also pursuing its largest-ever acquisition—a $7 billion all-stock deal announced in June to acquire Synaptics from EDA chip design software giant Synopsys. This move aims to capitalize on the growing demand in the AI device and robotics fields.

Management expects third-quarter revenue to be between $1.65 billion and $1.75 billion, with the midpoint of this range exceeding the average analyst estimate of $1.67 billion, according to LSEG data. CEO Hassane El-Khoury stated in the earnings release, "Our AI data center business remains our fastest-growing segment, and we now expect it to at least double in revenue by 2026, reflecting the strength of our intelligent power product portfolio and the increasing adoption of ON Semiconductor solutions across the entire power tree by our customers."

Second-quarter revenue for the period ending July 3 was $1.6 billion, a 9.2% year-over-year increase, slightly surpassing the consensus estimate of about $1.59 billion. Adjusted earnings per share (EPS) were $0.74, up about 40% year-over-year and above the market's expectation of $0.71. Management projects third-quarter adjusted EPS to be between $0.81 and $0.93, with the midpoint significantly higher than the consensus estimate of approximately $0.83.

ON Semiconductor continues to advance its "Fab Right" strategy to significantly cut costs and improve operational efficiency. In July, the company sold two manufacturing plants as part of this strategy. Other second-quarter financial highlights include: GAAP gross margin rose from 37.6% to 38.4%, and non-GAAP gross margin increased to 39.3%. GAAP operating margin improved from 13.2% to 16.1%, while non-GAAP operating margin rose from 17.3% to 20.8%. Net income attributable to common shareholders jumped from $170.3 million to $226.8 million, a 33.2% year-over-year increase. GAAP EPS rose from $0.41 to $0.56, a 36.6% gain. Operating cash flow surged 149.4% to $459.7 million, and free cash flow skyrocketed from $106.1 million to $425.4 million, roughly quadruple the year-ago period. The free cash flow margin expanded from about 7% to 27%. The company also repurchased $332 million in stock during the quarter. In other words, while revenue grew only about 9%, profit, cash flow, and shareholder returns expanded at a much faster rate, demonstrating that ON Semiconductor's Fab Right capacity optimization, cost discipline, and product portfolio upgrades are beginning to generate genuine operating leverage.

Segment data confirms that growth is increasingly concentrated in the data center power semiconductor business. The PSG segment revenue reached $829 million, up 18.7% year-over-year and 13% sequentially, now accounting for 51.7% of total revenue, up from 47.5% in the same period last year. The Analog and Mixed-Signal Group (AMG) posted revenue of about $545.7 million, down 1.8% year-over-year. The Intelligent Sensing Group (ISG) generated $228.8 million in revenue, a 6.6% increase. For the first half of the year, total revenue was $3.1168 billion, up 6.9% year-over-year. Adjusted EPS rose from $1.08 to $1.38, a 27.8% increase, and free cash flow was $642.6 million, up 14.6% year-over-year. These half-year and segment results indicate that ON Semiconductor has not yet seen synchronized prosperity across all its businesses. Instead, the PSG segment is accelerating first, driven by AI data centers, high-voltage power, and some automotive electrification demand, while AMG remains in a relatively moderate recovery phase.

The 'Chip Demand Craze' from the AI Infrastructure Boom Spreads from AI Chips and Memory to Analog and Power Semiconductors

ON Semiconductor's strong results show that AI chip demand is systematically spilling over from GPUs, ASICs, and HBM to power semiconductors, analog control, and monitoring signal chains. This is not simply analog chip companies "claiming the AI concept," but a physical necessity driven by increased computing density. ON Semiconductor has identified its AI data center business as its fastest-growing segment and expects related revenue to more than double by 2026. Texas Instruments saw its data center business revenue grow by about 90% year-over-year in the first quarter of 2026, with growth in the second quarter continuing to be led by industrial, data center, and automotive sectors. Infineon has joined the Nvidia MGX ecosystem, providing a complete solution from high-voltage conversion to GPU core power based on an 800V DC power supply, driving strong performance in the first half of the year.

The beneficiary scope of AI capital expenditure is therefore expanding from a few advanced process chips to a large number of relatively low-unit-price but high-volume, long-lifecycle, and reliability-certified basic analog devices for the data center power chain. The underlying logic for the strong demand expansion in the analog/power chip and power semiconductor/discrete device sectors, driven by the unprecedented AI wave, is that AI racks are leaping from tens of kilowatts for traditional servers to over 100 kilowatts, and evolving towards 600 kilowatts or even 1 megawatt. The next-generation GPU/TPU/ASIC single-chip power consumption could reach 2-4 kilowatts. Traditional 48V/54V architectures would generate thousands of amperes of current, leading to sharp increases in copper losses, heat generation, and busbar size. This is the driving force behind the global shift of AI data centers towards the ±400V or 800V DC architectures jointly promoted by ON Semiconductor and Nvidia.

Before electricity from the grid reaches the GPU, it must pass through AC/DC rectification, PSU, BBU, high-voltage distribution, an 800V to 50V intermediate bus conversion, and then multi-phase VRMs to reduce the voltage to the less than 1V needed by the GPU core. Each stage requires Si, SiC, or GaN power switches, gate drivers, digital controllers, power management ICs, and e-fuses. ON Semiconductor, Texas Instruments, and Infineon sell these "per-watt" devices. Therefore, the higher the rack power and the more complex the conversion stages, the greater the semiconductor value per cabinet. ON Semiconductor can provide EliteSiC MOSFETs/JFETs, high and low-voltage silicon MOSFETs, GaN devices, hot-swap smart e-fuses, multi-phase controllers, power stages, and PoL regulators across this "grid-to-GPU" power tree. This means a single AI rack not only needs more chips but also requires power devices with higher unit prices, higher voltage ratings, and greater efficiency.

The Nvidia MGX platform is a key strategic entry point for ON Semiconductor's AI data center business. The company already supplies power FETs, multi-phase power, SiC JFETs, and GaN solutions for existing MGX systems and directly serves Nvidia and suppliers of PSUs, BBUs, and future 800V power distribution boards within the MGX ecosystem. The standardized server and rack design of MGX also makes it easier for a certified power device to be adopted by multiple OEMs and ODMs.