The semiconductor map

The tariffs were the realignment everyone watched — loud, dated, tweeted. The chip controls are the one that will matter longer. Since 2018, and in escalating rounds through 2022, 2023, and 2024, the United States has walled China off from the most advanced semiconductors and, more importantly, from the tools to make them.¹ That isn't a tariff you can pay; it's a line drawn through the single most complex supply chain humans have ever built. And a supply chain that gets cut in half doesn't disappear — it reroutes. Vietnam is one of the places it rerouted to.

The slice Vietnam caught is the unglamorous one: the back end — packaging, assembly, and test, the final mile where a finished wafer becomes a chip you can solder. It is the back office of the silicon world. It is also, right now, one of the fastest-growing corners of the Vietnamese economy, and the question that decides how much it's worth is whether Vietnam stays the back office or climbs toward the front.

The chain, and where the wall sits

A modern chip is made in a global relay: it's designed in the US (and a few other places), etched by tools from a tiny handful of US, Dutch, and Japanese firms, fabricated in Taiwan, Korea, and increasingly Arizona, then shipped to the back end — packaging, assembly, test — before it reaches a device. The US export controls don't try to block the whole chain. They target the two chokepoints China can't easily replicate: the advanced tools and the leading-edge fabs. Cut those, and you've frozen a rival's ability to make the best chips without touching the rest.

Vietnam got the back office

Why Vietnam, for the back end? The same reasons it won the rest of China+1: proximity to the Chinese supply base, a young workforce, FDI-friendly policy, and a head start. Intel has run a packaging-and-test campus in Ho Chi Minh City for over a decade — it's the company's largest back-end site on earth, handling roughly half its global packaging and test volume.⁴ Amkor opened a $1.6B plant in Bac Ninh and, in early 2025, announced it would triple output from 1.2 to 3.6 billion units a year.³ Samsung keeps pouring in around a billion dollars a year. The whole Vietnamese semiconductor market is about $10B in 2025, on track toward $16B by 2030.²

The bet: from back office to front

The back end is good business, but it's the thin-margin end of a fat-margin industry. The value — and the strategic weight — climbs as you move up the stack from assembly to design to fabrication. Vietnam's explicit national bet, written into a 2024 strategy (Prime Minister's Decision 1018), is to climb: from packaging today toward chip design, and eventually some fabrication, by 2030 and beyond.⁵ And Washington, eager to make Vietnam a counterweight to China and to pull it out of the Chinese tech orbit, has every incentive to help it up the ladder.⁶

The transshipment trap, again

The same reflexive risk that shadows the whole China+1 story shadows this one, in a sharper form. The more Vietnam becomes the polite address for chips that are really Chinese — Chinese components packaged in Bac Ninh and stamped "Made in Vietnam" — the more it becomes a target for exactly the scrutiny the controls were meant to enforce.⁷ A back-end hub is, almost by definition, a relabeling hub; that's its commercial appeal and its political vulnerability at once. The success of the bet depends on Vietnam being seen as a genuine alternative to China rather than a backdoor for it — a distinction Washington will police, and one Hanoi can't fully control.

How to own the bet

Suppose you believe it — Vietnam's chip slice compounds and, maybe, climbs. The reflex, as always, is to buy a stock: a Vietnamese industrial-park name, or a foreign packager with Vietnam exposure. And, as we argued in the tariff piece, that reflex saddles you with basis risk — you wanted "Vietnam moves up the silicon stack" and you bought a company's management, leverage, currency, and governance instead. The cleaner instrument is a market on the proposition itself.

What's hard

The honest list is long. A leading-edge fab costs tens of billions and years, needs water, power, and thousands of specialized engineers Vietnam doesn't yet have — the strategy itself calls for tens of thousands of new ones. The export-control regime that helps Vietnam today could shift with an administration or a thaw. The transshipment risk above can turn an advantage into a liability overnight. And a market on "a fab by 2028" is exactly the kind of thin, long-dated contract that's hardest to trade well — few participants, wide spreads, capital locked for years.

None of that breaks the map; it scales it. The chip war isn't a headline that passes — it's a decade-long redrawing of where the world's most important objects get made, and Vietnam has a real, if junior, seat at the table. The back office is built. Whether it becomes the front is the bet — and the cleanest way to hold a view on it is an instrument shaped like the view, which is what we keep coming back to building at Seeker. The map is being redrawn now; the only question is whether you're positioned to be right about it on purpose.