Vietnam's Ministry of Information and Communications published an updated five-year semiconductor industry plan on 12 May 2026, signing off on US$25 billion in committed foreign direct investment by 2030. The number is meaningful but the breakdown matters more: roughly US$3 billion is allocated to greenfield wafer fabrication, and US$18 billion to OSAT (outsourced semiconductor assembly and test). The remaining US$4 billion is design centres, training infrastructure, and supplier facilities.
Why the OSAT-heavy mix matters
The Vietnam plan is, in industrial terms, a packaging and test strategy, not a fabrication strategy. OSAT operations require less capital intensity per facility (typically US$500m-US$1.5bn versus US$8-20bn for a leading-edge fab), shorter ramp times (18-30 months versus 4-7 years), and substantially lower entry barriers in terms of skilled workforce density. They also produce different export revenue patterns — packaging margins are around 12-18% versus 35-45% for leading-edge fab output.
The strategic logic is sober. Vietnam cannot compete with Taiwan, Korea or even China on advanced fab capacity within a five-year horizon — the supplier ecosystem, talent base and capital depth simply do not exist. Where it can compete, and is competing, is in the OSAT layer that has been migrating out of China since 2022 as US-China decoupling has accelerated.
The foreign partner profile
The biggest single committed investor is Amkor Technology, with US$1.6 billion across two facilities near Bac Ninh (one already operating, the second under construction with first output expected Q4 2026). Intel's Saigon Hi-Tech Park expansion, the second-phase US$1.2 billion announced in 2024, is now in late-stage equipment installation with first output for 2027. Samsung's Thai Nguyen complex is adding US$0.8 billion of packaging capacity. ASE (Taiwan-based) is in negotiations for a US$1.4 billion site near Hanoi.
Notably, no leading Chinese OSAT player is on the roadmap. The exclusion is deliberate — both because of US technology-transfer concerns flagged in the bilateral Vietnam-US framework agreement signed in 2024, and because of Vietnam's own bilateral commercial preferences in the current geopolitical alignment.
The training pipeline that everyone underestimates
The single most-cited constraint by every foreign investor in the roadmap is engineering talent. Vietnam has approximately 5,000 semiconductor-trained engineers; the industry plan calls for 50,000 by 2030. The expansion requires accredited engineering programmes at five universities (currently only Hanoi Polytechnic has a credible specialised programme), specific government-industry training partnerships, and active foreign recruitment from the Vietnamese diaspora in Taiwan, Korea and California.
The May 2026 update commits Vietnam to direct subsidies for returning engineers — a US$30,000 relocation bonus and a tax holiday on Vietnamese income for the first five years, designed specifically to draw mid-career Vietnamese engineers from TSMC, Samsung and Intel back to domestic operations.
What this means for the regional value chain
If the plan executes at even 70% of stated targets, Vietnam will be the largest OSAT cluster outside Taiwan and China by 2030 — likely overtaking Malaysia, which has been the regional packaging hub since the 1990s. The Malaysian incumbent industry (in Penang) is responding with its own US$8 billion industry support package announced in March 2026, but the demographic and labour-cost advantages tilt toward Vietnam in the medium term.
For investors, the practical signal is that the Vietnam-semiconductor trade is no longer a 5-10 year speculative thesis. The capital is committed, the partners are named, and the facilities are physically under construction. The remaining questions are execution velocity and the engineering talent pipeline — both of which are visible and measurable on a quarterly basis.