Does it being very high precision or whatever mean that it must cost $75M per unit? Is it made of gold?

The cost primarily comes from the work needed to be done to make sure it works, not the raw materials. But there probably are some pretty expensive materials in it too.

https://en.wikipedia.org/wiki/Advanced_Inertial_Reference_Sp...

This is what we had to build in the 60s to allow a missile to know where it physically existed precisely enough to allow it to 50% of the time hit within a circle of ~50 meters.

When you get to certain points in physics, certain energy regimes, you no longer are building machines or tools or something mass market. You are building artisan scientific instruments, and then sometimes gluing explosives to them.

Even modern laser ring gyros do not even share a dinner table with the precision and accuracy of the above singular component of the Peacekeeper ICBMs, and that was a long time ago.

"Tech" and some of the developments of the past few decades have really confused people. The miniaturization of the transistor, and building billions of transistors on a small slice of silicon is an aberration, an anomaly. Most things don't get "Better and better and cheaper and cheaper" like that because shit just doesn't scale infinitely and in general materials science isn't that precise.

For these ground based midcourse interceptors, they have to precisely loft the interceptor package at an incoming projectile. They have to shoot a bullet with a bullet, except the target bullet might even be moving around a bit, and your ability to precisely quantify the exact parameters of it's position and velocity is already limited. Is your position and velocity measurement an inch off? Two inches? Is that too much?

How well have you quantified the thrust of your rocket engine? THIS specific rocket engine, not a random one from the batch. Will you be off in a direction by a few meters per second? That might be enough to scuttle your interception.

IIRC this is the interception payload, a kinetic kill vehicle:

https://www.youtube.com/watch?v=KBMU6l6GsdM

Then, are any of those nozzles slightly less smooth than they should be? That's a miss. Did propellant slosh in an unexpected way? That's a miss, Chicago is now a smoldering crater. Incoming round bounce your rader signal in just a slightly different way than you had the data to know about? Miss, San Fransisco now has significantly cheaper real estate. Chaotic properties of hot expanding gas slightly different than your simulation in the unluckiest way? Miss.

You have to exhaustively inspect, reinspect, quality control, test, simulate, retest, catalogue, document, every single component. You have to be able to predict, almost perfectly, how every single component will act and perform in a situation you will never get a test for.

High energy physics is always going to be hard, never cheap, because high concentrations of energy are literally what the universe itself is trying to reduce. The rules of reality itself are against you.

A creator on youtube named Alexander the Ok has done wonderful videos on a lot of the technology that goes into these systems, especially older, less classified systems.