From Newsgroup: rec.aviation.military
"Jim Wilkins" wrote in message news:10mm61v$20kef$
1@dont-email.me...
"Nick Charles" wrote in message news:
R46cnU_XEfW8wRP0nZ2dnZfqn_qdnZ2d@supernews.com...
Thus, no one in 1965 foresaw everyone having a battery powered, wirelessly >networked, cool, quiet and inexpensive computer in their pocket. Because
it was physically impossible!
No one included Gene Roddenberry and Stanley Kubrick. Yet the Intel 4004 arrived in 1971.
-----------------------------------
The technology that enabled a microcomputer had been developed in 1959,
though only its developers realized its potential.
https://www.nutsvolts.com/magazine/article/the-birth-of-the-integrated-circuit "Although the military was informed of this progress, they were not enthusiastic."
The military wants to see it practical, they have been repeatedly burned by promising ideas that turned out poorly, such as cramming turbochargers into sleek fighter planes. The machine gun needed 40 years to become light enough for one man to carry. Governments are the only customers willing and able to provide early development funds and wait for the product so they led the advance of much technology.
Much of the difficulty was in the microscopic fabrication process which was solved by hands-on technicians as well as engineers. An example is the photographic process that hides text in a period-sized dot for spies. I've read that the source of sufficiently fine grained emulsion was dimensionally stable glass plates for astronomical photography. I developed a
spectrographic image on one and measured the positions of the extremely fine lines it captured with a microscope. The individual black amorphous silver grains were quite visible.
https://en.wikipedia.org/wiki/Microdot
"The reduction was such that a page of text would be legibly reproduced in
an area of 0.01 mm2."
I was closely involved with electronic testing during the 70's and 80's. The limit to progress was the fault rate, the percentage yield of good product. Fabrication advances immediately went to making more difficult product. I
have a stack of rejected 6" Silicon IC wafers mostly covered with the ink
dots that marked defects.
This is what IC wafers look like. Each identical square is a separate device like the four shown magnified behind the round wafer.
https://chipscapes.com/products/silicon-wafer-with-microprocessor-chips-4-inch-aloha
Typically all grades of a device are made the same way and testing sorts
them into higher or lower speed or functionality. Radio Shack sold the ones that worked but didn't quite pass spec. I find that on components from
Amazon too, a batch of surge voltage clamps tested either slightly above or slightly below the tolerance limits, none within them. Much component
testing can be done using lab power supplies with meters and adjustable current limits.
My training to become an IC manufacturer's lab tech was reverse engineering
a packaged IC back to the schematic by etching it down layer by layer and tracing the connections. I got the FETs, bipolar transistors, current
mirrors, logic gates, op amps and comparators right, all but the voltage references which were jumbles of junctions with no contrast between them. As the lab tech I applied what I'd learned of technical photography to make current flow visible.
--- Synchronet 3.21b-Linux NewsLink 1.2