I hated to leave it at just my own general feeling on prices if I could find real data, so I went searching. I didn't find the charts I wanted for tubes but I did for transistors.
The price of a transistor has had an amazing ride. See the chart below.
The prices are listed in billionths of a dollar, that is ten to the minus ninth power. The price curve is in black, and refers to the right hand scale. So the picture says that in 1955, there were something like ten transistors made, and they each cost on the order of tens of dollars each. A few years later, production jumped to hundreds and price fell to between dollars and dimes. Shortly afterwards, the IC was invented and transistor processing really took off. The classic illustration is that sometime in 2010 the cost of a transistor in an IC was less than the cost of a single letter of newsprint.
So transistors started off as lab curiousities, and then the price headed for the floor. Tubes were a fully mature mechanical technology in terms of manufacturing. So the cost of manufacturing a tube was essentially constant, even increasing as manufacturing facilities went off-line.
Well, technically, it was the joining of formal descriptions of binary logic with first relays, then vacuum tube logic circuits, and finally the path through diode logic, RTL, TTL, ECL, and finally CMOS logic that made the moon missions computable. Relay based calculators and calculating/sorting machines were available commercially back in the first couple of decades of the 1900s. In fact, mechanical computers were invented, just not completely made practical and general purpose were available in the late 1800s after their invention by Herman Hollerith. He founded one of three companies that would merge to become IBM eventually.
But before that, Ada Lovelace worked with Charles Babbage in the early-mid 1830s to develop Babbages ideas of the Analytical Engine into a usable computer.
The concepts of storing data and programs in a digital manner on mechanical, then electronic machines resulted in vacuum tube computers in the 1940s and 1950s. These things were a mess to keep running. The entire computer industry heaved a sigh of relief when they got transistors and then those funny four-transistors-on-a-chip that underlaid the whole generation of things like IBM's SLT - Solid Logic Technology.
The opamp was used in analog computing schemes in WW2. It was the basis of certain bombsights and the aiming computers for battleship guns. These were vacuum tube based. But as soon as we got transistors and learned to use them, we went right after implementing linear analog computing with transistors too.
The solid state opamp was much like the solid state computer - an application of newfound small, low power devices to already known kinds of devices. So I'd say that modern opamps and modern computers were siblings, not one the precursor to the other.
I worked for a guy who was a programmer for the minicomputers used on the moonshots back when. He had some very interesting stories, one of which included the fact that on certain moon shots, certain program combinations and operations in certain memory locations hung the computer. Not what you want when you're landing or doing a break-orbit burn. They solved the problem by rewriting the programs to NEVER use those areas of memory, and the shots were successful.
I understand the Russians used a device encompassing a large cylinder with pins in it that was spun at a slow, constant rate to do all the critical operations and timing - yep, a variant of the pin-type music box. Seems laughable until you realize it was indeed a computer, just a mechanically programmed one with limited flexibility.
"It's not what we don't know that gets us in trouble. It's what we know for sure that just ain't so"
Mark Twain
R.G. wrote: ↑Mon May 21, 2018 2:32 am
Well, technically, it was the joining of formal descriptions of binary logic with first relays, then vacuum tube logic circuits, and finally the path through diode logic, RTL, TTL, ECL, and finally CMOS logic that made the moon missions computable. Relay based calculators and calculating/sorting machines were available commercially back in the first couple of decades of the 1900s. In fact, mechanical computers were invented, just not completely made practical and general purpose were available in the late 1800s after their invention by Herman Hollerith. He founded one of three companies that would merge to become IBM eventually...
RG - Wonderful computer and computational history lesson!
When I was in high school (graduated 1980), I was in BOEC and OEA clubs that did competitive testing on such knowledge. Well, just the computer stuff (current at the time and historical). So, we did (nerdy) things like learning the Hollerith coding system that was used on his classic punched cards. Which served the computer world at least in to the 80's (amazing in it's own right). When I was a Computer Science major at the University of Michigan, most of my programming was done on punched cards. So, programs were stored as a deck of punched cards - and to run them, we would submit a batch job through a card reader - to load the program from cards and run - sending output/results to a line printer. Coincidentally, that was the same time that IBM decided to get in to the PC game. Which legitimized PCs for commercial use. Of course that revolution would have come anyway - but IBM certainly hastened the process by a bunch of years. Which in a sense begot Microsoft. As, IBM chose to have Digital Research write their operating system (PC-DOS). And Bill Gates was cleaver (lucky?) enough to buy the rights from DR. This was the goose that laid the golden egg for MS. Which provided an enormous cash stream with which to build a software empire.
