You run a chip company for forty years. A fab takes four years to build. A chip takes three years to design. By the time any decision pays off, the world that justified it is gone.
Every lever has a delay fuse welded to it. You never act on now — you act on a quarter years away, using information you don't have yet.
The main screen is dominated by this strip. The strategic act of the game is reading its shape against the demand forecast — and noticing they don't match. Notice the fuses you can't pull back.
Building a fab feels safest at the top of the cycle — cash is abundant, customers are begging, the board is thrilled. It arrives twelve quarters later, into a glut, with depreciation attached.
The sim rewards neither hoarding nor building. Starve your customers and they design you out permanently; overbuild and depreciation eats you in the trough. Only the timing saves you — and the timing has to be decided four years blind.
Wafer lots are guests. Tools are rides. A queue is a queue. A tool down is a broken ride with a mechanic walking over — except it costs $2M an hour and nobody notices for a quarter.
You don't need a report to know your implanter is the constraint — you can see the line. That's the entire argument for the spatial layer.
RCT told you the truth about your park through guest thoughts. Here, engineers do it — and your management style can switch them off. Flip the toggle and watch what happens to your early-warning system.
Under LOW, the bubbles don't turn angry — they thin out and go quiet. New players will read that as things going well. Six quarters later the errata lands and there is no record of anyone warning you. It costs nothing to implement and it's the scariest thing in the game.
Every decision silently writes a face-down card into a future quarter. You see its shape long before its content — a grey marker on the calendar, a maintenance log slipping, a thought bubble repeating.
This makes boring maintenance work mechanically correct, which no tycoon game has ever managed. Good decisions write cards too — the engineer you backed in 2029 invents your packaging advantage in 2036.
Mouse-wheel zoom moves between them continuously. Time runs continuously too — pause, 1×, 2×, 4×, with auto-pause on anything that matters. Not turn-based.
Weighting: the campus and the floor are where you live. The world map is a logistics and intelligence board you check, not a 4X map you play.
Civ's adversarial nations, but you never leave your campus. They have to reach you where you live: a recruiter's car in your lot, a billboard going up across the street, your supplier's output pre-bought out from under you.
No diplomacy menu, no war declarations. Just four opponents with legible doctrines, a forty-year arc, and the growing sense that things are being done to you on purpose.
Each one is a decision with a lead time and a ripple. Swipe through.
Not ten margin numbers — ten personalities. You don't sell chips, you win sockets: a two-year sales cycle that pays out for eight years. Your revenue today was decided in 2019.
Cheapest kill-switches first. Two beloved references can produce a game that does neither well — the schedule is built to find that out early.
The honest risk: step 5. If zooming between the floor and the rivalry feels like two games stapled together, we need to know in month three, not month thirty.