[throwawaygh]

No! It's negative WRT chip production rate! I have a phd in controls and meant what I said; I assume roboticsresearcher also knows some freshman-level control theory ;-).

Stability ==== good in undergrad engineering, but not here. We DON'T want production rate to be stable when we have a global supply shortage! Here, a negative feedback loop is stabilizing the system in an undesirable equilibrium.

I.e., the function that's being controlled in "supply of chips", the stable state is "saturated supply", and the negative feedback loop that maintains that equilibrium is "starving chip fab suppliers".

(meta: people down-voting comments on control theory terminology by two different experts in this field at least makes me feel a bit better about the signal:noise ratio on the vote counts on my other comments in this thread ;))

[saltcured]

Only partially facetious here: you have mistakenly decided that it it clamps at zero. If we consider the hoarding aspects, the feedback continues into net negative chip availability. Before long, there will be roving gangs of looters taking back the chips you thought you already had! ;-)

[throwawaygh]

Yes, it's fair, I do assume that chip supply/demand isn't subject to the forces of roving gangs of looters :)

[QuarterReptile]

That's fair. I was looking at it wrong.

[alisonkisk]

It's not stabilizing if the economy is in a death spiral of failing interdependent companies who can't produce machines for each other.

[robotresearcher]

In control terms it is. There’s a stable asymptote of ‘we made no machines today’ every day, until some external process breaks the dependency cycle.