SK hynix approved about 54 trillion won of new fabs on 7 August 2026, the largest single item in a Korean memory build-out that domestic coverage now prices in thousands of trillions of won. We went looking for the announcements that carry both an amount and a date. Four exist, all from one company, and they total 85.3 trillion won. The earliest cleanroom among them opens in February 2027, and the largest two open in December 2028 and June 2029.
Last verified: 9 August 2026. Built from SK hynix newsroom releases dated 3 April 2024, 25 February 2026 and 7 August 2026, plus a Korean brokerage sector report we could not open and Korean market reporting where no company document exists. The dated commitments should hold. The programme totals circulating in the press have already moved twice and will move again.
Editor’s opinion
The calendar is the story here, not the size
Our view is that the useful number in Korea’s memory expansion is not the announced total but the schedule underneath it, and every date SK hynix has published points past 2027. The company’s own end-to-end example is Indiana: an investment agreement signed in April 2024, mass production stated for the second half of 2028. Four and a half years, on a packaging site rather than a full wafer fab.
The counter-argument is that acceleration is itself the signal, and it is a fair one. SK hynix moved its first Yongin cleanroom from May to February 2027, and the 21.6 trillion won it committed in February 2026 had been an industry estimate before it became a company release. Announced numbers do become filed ones. Our view survives because each acceleration so far has moved a building date rather than an output date, and because the company has not restated when any of this becomes wafers. It would not survive a release that gives a production start earlier than 2028.
Andrew Lee. This section argues a view rather than establishing a fact. It is not a forecast and not a recommendation, and the material it draws on is listed in the sources below.
What is committed, with an amount and a date
Four SK hynix announcements between April 2024 and August 2026 carry both figures. Together they commit 85.3 trillion won inside Korea and 3.87 billion dollars in Indiana. We have left the dollar figure unconverted, because turning it into won would import an exchange rate assumption that the company never made.
| Announced | What was committed | Amount | Company’s stated target |
|---|---|---|---|
| 3 Apr 2024 | Advanced packaging fab, West Lafayette, Indiana | $3.87bn | Mass production, second half of 2028 |
| Jul 2024 | Yongin first fab, initial facility investment | 9.4tn won | Not restated in the February 2026 release |
| 25 Feb 2026 | Yongin first fab, phases 2 to 6, five further cleanrooms | 21.6tn won | All cleanrooms by end December 2030, first cleanroom February 2027 |
| 7 Aug 2026 | Yongin Y2 | 35.2tn won | Cleanroom operating June 2029 |
| 7 Aug 2026 | Cheongju M17 | 19.1tn won | Cleanroom operating December 2028 |
The February 2026 release is the one worth reading closely, because it does the arithmetic itself: 21.6 trillion won on top of the 9.4 trillion announced in July 2024 brings the first Yongin fab to roughly 31 trillion. It also describes what has physically gone up, a fab frame the height of a fifty-storey apartment block. The August decision then adds a second Yongin fab and a Cheongju fab beside it. If you want the twelve-year engineering record that sits behind that August board vote, we went through it in SK hynix did not win HBM in 2023.
A cleanroom is not a wafer
Every target in the table above is a cleanroom date, and a cleanroom opening is the handover of an empty room. What you get on that day is a shell, a floor and filtered air. The machines that will fill it cost more than the building does and have not arrived.
SK hynix says this explicitly. The February 2026 release ends its investment figure with a parenthesis: equipment installation costs are not included. So the roughly 31 trillion won attached to the first Yongin fab buys two building shells and six cleanrooms, not a running production line.
The company has published exactly one lead time that runs all the way to output, and it is the Indiana one: agreement April 2024, mass production second half of 2028. Hold that against the Korean dates and the picture is straightforward. A cleanroom opening in February 2027 is the front of a queue, not the end of one, and the Cheongju and Yongin rooms open a further twenty-two and twenty-eight months after it. Nothing in the documents supports meaningful new output from these sites before 2028.
This does not mean Korean memory supply is flat until then. Capacity added inside fabs that already exist is a separate question, it moves on a much shorter cycle, and this article does not cover it. The claim here is narrower and it is about the new sites only: the money announced in the 2024 to 2026 wave is buying rooms that are still empty.
Samsung is half the story and almost none of the record
Samsung Electronics is the larger builder of the two, and it is the weaker half of this article. We could not find a Samsung-published document that gives a figure and a date for the Pyeongtaek P5 and P6 phases or for its plot in the Yongin national industrial complex. What exists is Korean media reporting of a June 2026 announcement, and a brokerage sector report summarising it.
That is not enough for us to print a number. The totals attributed to the same Samsung programme differ between the sources we saw, which is the clearest possible signal that they are being assembled by third parties rather than disclosed by the company. When two accounts of one announcement disagree, printing either one is a coin toss dressed as reporting.
Absence of a filing is not absence of a plan, and it would be wrong to read this as Samsung standing still. Pyeongtaek has been under near-continuous construction for a decade and P4 is a real building. The gap is in what we can cite. If Samsung does file the numbers, the place they will appear is DART, and you can read that yourself without Korean: we set out how in reading Korean corporate filings in English.
The programme totals we are not printing
The figures that make these projects sound like national infrastructure are the cluster-level ones, and they are the figures we are withholding. The last Yongin cluster total SK hynix published itself was 120 trillion won, in the April 2024 Indiana release. In February 2026 the company said the projected scale would increase significantly against that level, citing an eased floor area ratio under Korea’s high-tech strategic industry act and inflation, and it gave no replacement number.
Five figures in this article
Where each one stands, as at 9 August 2026
| Figure | Attributed to | Status |
|---|---|---|
| 85.3 trillion won across four dated commitments | SK hynix newsroom, February 2026 and August 2026 | On the record, company-published |
| 120 trillion won for the Yongin cluster | SK hynix newsroom, 3 April 2024 | On the record, but the company says it has since risen |
| Current Yongin cluster total | Trade press, industry estimate | Not confirmed — figure withheld |
| Samsung cluster investment totals | Korean media and a brokerage report | Not confirmed — figure withheld |
| Fab-by-fab equipment schedule for 2026 and 2027 | Shinhan Securities sector report, 6 August 2026 | Cited, original not read |
The first two rows are the company reporting its own decisions, which is the correct source for both. The third and fourth are assembled by others from statements we cannot open, so we print neither. The fifth is a paid report distributed through a brokerage platform with no public address; we name it because it is where much of the circulating detail originates, and we reproduce none of it.
One date to check before you trust a circulating timeline
The sector report is dated 6 August 2026, the day before the 54 trillion won board decision. Any fab-by-fab timeline derived from it therefore predates the largest commitment described in this article.
If you are trying to size this build-out, the honest answer is that the size is not yet documented and the schedule is. That asymmetry is unusual and it is worth noticing. Korean chipmakers have historically disclosed capital plans in round programme totals and stayed vague on timing; here the timing is precise and repeated in company releases while the totals have been left to others.
None of this is a view on either share price. It is a statement about when concrete becomes capacity, and on the documents that answer is 2028 at the earliest.
Sources
We take facts from these and write our own sentences. Where a figure originates in a company filing we say so, and where it originates in a research house estimate we name the house. We link to a document only where we hold the exact address for it; a link to a publisher’s front page would not get you to the source, so where we do not have the direct address we give you enough to find it.
Company announcements
- SK hynix Newsroom, New Facility Investment for Yongin Semiconductor Cluster, 25 February 2026. Source for the 21.6 trillion won commitment to end December 2030, the 9.4 trillion won announced in July 2024, the resulting total of about 31 trillion won, the two shells and six cleanrooms, the first cleanroom moving from May to February 2027, the exclusion of equipment installation costs, and the statement that the projected cluster scale will rise significantly.
- SK hynix Newsroom, SK hynix Signs Investment Agreement of Advanced Chip Packaging with Indiana, 3 April 2024. Source for the 3.87 billion dollar Indiana investment, mass production in the second half of 2028, and the 120 trillion won Yongin cluster figure.
- SK hynix Newsroom, SK hynix Invests 54 Trillion Won in Yongin Y2 and Cheongju M17, 7 August 2026. Source for the 35.2 trillion and 19.1 trillion won split and the June 2029 and December 2028 cleanroom targets.
Material we could not open
- Shinhan Securities, ‘Semiconductor equipment — CapEx Speed Up!’ sector report, 6 August 2026, by Park Hyun-woo and Kim Ye-seong. Distributed through the firm’s research platform with no public address, and we have not read it. Named because it is the origin of the fab-by-fab schedule and the programme totals now circulating in Korean coverage. We reproduce neither.
- Samsung Electronics: we found no company-published document setting out an amount and a date for the Pyeongtaek P5 and P6 phases or for the Yongin national industrial complex. Listed here so the gap is visible rather than inferred from our silence.
- Korean market reporting, June and July 2026, on Samsung regional investment plans and on the current Yongin cluster total. Read as evidence of how the programme is being discussed, not as evidence of the amounts.
Corrections
None yet. If you can point to a primary document that contradicts anything above, we will correct it and say what changed.
Disclosure. The author holds no position in Samsung Electronics or SK hynix as of the date of publication. This article is general information and analysis about announced capital expenditure, construction schedules and company disclosure. It is not investment advice, not a recommendation to buy or sell any security, and not a forecast of any price or index level. The section headed Editor’s opinion is comment: it argues a view, and we fence it off from the sourced material for that reason. We have not tailored anything here to your circumstances, and only a qualified adviser in your jurisdiction can do that. We are not registered as an investment adviser in any jurisdiction, and this site has no commercial relationship with any company, exchange or research provider named above. See the Editorial & Sourcing Policy and Disclaimer.