Chuck Guzis wrote:
...
Nowadays many "hardware" positions seem to rely on one's ability to
spew Verilog or VHDL and not know which end of the soldiering iron is
the hot one.
--Chuck
Chuck, that was spoken like a true old-timer, but your derision is
laughable. "Hey, you damn verilog coders, get off my lawn!"
I'm surprised, as you saw the transition from tubes to transistors to
ICs, so it shouldn't shock you that the next evolution of integration
requires just as much engineering and intelligence, although the
problems to be solved are somewhat different.
And believe me, even with hardware jobs consisting of "spewing" verilog
code, there is quite a difference in the work of a hardware job and a
software job. Some can do either job well, but not many.
No doubt there are challenging designs that are done at board level
(e.g., DRAM interfaces with a 300ps data cycle time), but most of it is
simply connecting together the chips in a prescribed way, kind of like
doing a connect the dots drawing. The action these days is inside the chip.