AI power demand drives record growth
- Q2 2026 revenue was a record $574 million, aided by broad strength across major segments.
- Non-GAAP gross margin reached 41.9%, keeping the long-term 43% goal clearly in view.
- Management raised the 2026 Data Center growth target to at least 50%.
- Semiconductor revenue hit a record $278 million in Q2, up 33% year over year.
- The main worry is concentration, as one large data center customer can move the entire story.
Two growth engines fire together
AEIS sells power systems that sit inside high-value equipment. The current story is clear: AI data centers need more power, and chip equipment demand is recovering at the same time. In Q2 2026, total revenue reached a record $574 million, and non-GAAP EPS climbed to $2.74.
The bull case strengthened significantly after Q2. Management upgraded the full-year 2026 Data Center growth expectation to at least 50% and projected nearly 50% year-over-year growth for second-half Semiconductor revenue. With a new Thailand factory expected to deliver its first revenue in Q4 2026, the company is building a global footprint capable of supporting $5 billion in total revenue.
Margins are central to this story. AEIS achieved a 41.9% gross margin in Q2. The next proof point is whether the company can maintain levels above 41% consistently without one-time tariff refunds, especially while aggressively adding capacity in places like Thailand.
The bear case revolves around outside shocks and concentration. A sharp slowdown in AI data center spending, a pause by a major hyperscale customer, or a severe macroeconomic downturn could quickly cool the growth trajectory. The overall Finn score is positive but balanced, highlighting the need to watch valuation and customer concentration.
Power hardware built into complex machines
Advanced Energy designs and sells precision power conversion, measurement, and control products. Its systems help control electricity in complex machines, such as semiconductor tools, AI servers, medical devices, and industrial equipment.
The company makes money by selling these products directly to original equipment makers, often after lengthy design work with the customer. That design process creates sticky relationships because changing a power system inside a complex machine is difficult and carries high risk.
The model works best when AEIS wins designs early, and the customer then ramps production. That is happening right now in AI data centers and semiconductor equipment. It can break when demand cycles turn down, when one large customer changes its plans, or when new factory ramps hurt delivery times or weigh on margins.
Where the power systems go
Semiconductor power
RF generators, matching networks, and plasma controls help chip tools run critical etch and deposition steps. New eVoS, eVerest, and NavX products are driving market share gains.
Data center power
High-power shelves and DC-DC modules serve AI servers and hyperscale systems. This is a rapidly growing area, with management raising its 2026 growth target to at least 50%.
Industrial power
These platforms serve uses like glass coating, test and measurement, and battery production. The market completed inventory rebalancing and is showing normalized demand.
Medical power
Medical products support diagnostic and therapeutic equipment. This market relies on steady, recurring capital spending cycles.
Telecom and networking power
Telecom and networking systems are smaller today, but AI-related networking programs are adding demand, growing 12% year over year in Q2 2026.
Next-generation high-voltage power
AEIS is developing 800-volt solutions for future data center power designs. High-volume production is expected in 2028.
Q2 mix: semiconductor and AI lead
Segment mix is based on Q2 2026 revenue: Semiconductor $278 million, Data Center Computing $192 million, Industrial & Medical $80 million, and Telecom & Networking $24 million. The data center segment is growing fast but carries customer concentration risk.
What could break the setup
A large data center customer slows orders
High impact · Medium oddsData Center Computing is a major growth driver, and the segment is highly concentrated. If one large hyperscale customer changes timing, shifts suppliers, or faces its own limits, AEIS could miss its 50% growth target.
Semiconductor cycle turns again
High impact · Medium oddsSemiconductor revenue was a record $278 million in Q2 2026. This market is highly cyclical. A cut in chip equipment spending would hit the company's largest business segment.
Capacity ramps strain margins
Medium impact · Medium oddsAEIS reached a 41.9% gross margin in Q2, but it is expanding capacity heavily to meet demand. The new facility in Thailand must ramp cleanly. Delays or higher labor costs could hold margins below the long-term 43% goal.
Industrial and medical demand falters
Medium impact · Low oddsIndustrial & Medical revenue was $80 million in Q2 2026, showing normalized demand after an inventory rebalancing phase. The risk is that macroeconomic weakness reverses this recent recovery.
Tariffs pressure product costs
Medium impact · Low oddsTariffs and trade policy remain a background risk. While AEIS benefited from some tariff refunds in Q2, future higher input costs or sudden supply shifts could pressure gross margin.
In one breath
What does Advanced Energy Industries do?
Advanced Energy makes precision power systems. Its products help control electricity in chipmaking tools, AI servers, industrial equipment, medical devices, and telecom networks.
Why is AEIS tied to AI?
AI servers need high-efficiency power hardware. AEIS sells high-power shelves and DC-DC modules for data center systems, and its Data Center Computing revenue grew to $192 million in Q2 2026.
Is AEIS mainly a semiconductor company?
Semiconductor is its largest segment by Q2 2026 revenue, but Data Center Computing is close behind and growing faster. The company also sells into the Industrial & Medical space.
What is the biggest risk for AEIS stock?
The biggest company-specific risk is customer concentration in Data Center Computing. If a major AI data center customer slows spending or changes suppliers, growth could cool quickly.
