Quanta Raises 2026 Capex to NT$40 Billion as AI Server Demand Visibility Reaches 2028

The available reports also do not identify the customers behind the 2028 outlook, distinguish binding orders from planning visibility, or state the volume or revenue attached to that horizon. The forecast is a management signal supporting the investment case, not a publicly measurable order backlog.

The Taiwan step is a pending factory transaction

The longer-dated part of the plan is Quanta’s proposed NT$19.7 billion acquisition of AUO’s Huaya plant in Guishan, Taoyuan. The companies were still negotiating contract details at the time of the investor conference. Quanta expects a fourth-quarter 2027 handover and says bringing the facility into operation in 2028 could double its Taiwan AI-server back-end assembly capacity from the current level.

The attraction is specific to this site, not merely added floor space. The former panel plant is in the same Huaya technology park as Quanta’s research-and-development headquarters and existing Taiwan server plants, a configuration the company says would connect design through final production more efficiently. Its higher electrical-capacity configuration matters because high-end servers require extensive pre-shipment testing. The report describes the facility as a way to obtain power-ready capacity faster than building it from scratch.

The two capacity claims should not be merged. The end-2026 objective concerns AI-related capacity relative to the end of 2025; the 2028 forecast concerns Taiwan back-end assembly relative to its current level. Nor is the Huaya transaction final, so it cannot yet be treated as delivered capacity.

Operating scale and funding are separate questions

Quanta’s scale gives context to the factory push. It reported second-quarter revenue of NT$1.04 trillion and net income after tax of NT$28.651 billion, up 69.9% year on year, with earnings per share of NT$7.43; the same account says each was a quarterly record. First-half net income was NT$49.844 billion and first-half earnings per share NT$12.93. Those results describe current operating performance, not the return that future factory spending will earn.

The company also recently completed a US$2.2 billion global depositary receipt offering, described as one of Taiwan’s largest recent international equity raises. It has separately approved a US$973 million capital increase for its U.S. subsidiary QMN to add California AI-server assembly capacity, while a €19.5 million increase for German subsidiary QCG is for intelligent-driving automotive production. The U.S. and German outlays are to be made in stages based on operating conditions and construction progress. The company’s disclosure does not establish whether those capital increases or the GDR proceeds are included in, or finance, the NT$40 billion capex target.

That separation matters. Adding the factory purchase, capex budget and subsidiary funding rounds into one total would double-count or imply a financing relationship the disclosures do not provide.

What will resolve the 2028 bet

The next meaningful evidence is concrete: a signed Huaya agreement, the planned late-2027 handover, and whether Quanta meets its end-2026 capacity target while maintaining the margin level it forecasts. The company has also said AI-server revenue share should rise in the second half, which will offer a nearer-term test of the product-mix direction behind the expansion.

Until Quanta identifies the scale and durability of the orders underpinning its 2028 visibility—or reports how the Huaya transaction and overseas subsidiary funding fit with the capex budget—the public record supports a large, partly executed manufacturing plan, not a verified measure of the demand that will ultimately fill it.

PwC finds rising senior-skill demands in AI-exposed junior jobs

PwC finds growth in a subset of AI-exposed junior job postings demanding senior skills, alongside pressure on entry-level opportunities. Workplace research suggests AI can support learning, but leaves employers to demonstrate how recruits acquire independent expertise.