This is the part I don't understand about chip export restrictions; the argument used is that high-end chips would be used for high-tech weaponry, but most weapons don't need high-end chips, 2-5 generations old chips are good enough for most smart weapons. I mean the Tomahawk is from '83, the HIMARS from the late 90's, etc. I don't believe that high end, small process chips are used in any weapons right now.
The only use case is modern day AI workloads, but that's used more in planning than in the field. I can imagine a use case for e.g. image recognition, but again, that tech or the level required is not new at all and doesn't need state of the art chip tech.
The only use case is modern day AI workloads, but that's used more in planning than in the field. I can imagine a use case for e.g. image recognition, but again, that tech or the level required is not new at all and doesn't need state of the art chip tech.