Regardless of the academic dispute of splitting the discipline into computer science and engineering - a valid but moot point- once one enters the field of applied informatics such as process control, data acquisition and programmable controllers of any kind, I believe electronics remains a fundamental pylon . In this very physical layer of industrial, immediate application all abstraction programmers are accustomed to evaporates to leave behind direct hardware related calls, and that leads to the need for a good understanding of the basics.System engineers need to know that perfect system they just designed has sensor-related tolerances and non-linearity, needs compensation of different types and degrees and linearizing circuits have their limits, that look-up data tables just make for an easier implementation of a very computing-intensive equation on low power distributed controllers, etc.Even when considering a new digital design for a new type of measuring device, using some sort of sensor, questions of how fast the system needs be, what bandwidth etc- find relevant answers in basic electronics not in digital design. While it is simple to maximize everything sometimes low-cost, low part number designs fully cover the possibilities of the sensor- and help reduce cost, and unnecessary overhead in more complex systems they will be integrated into.
My opinion is that someone neglecting basic electronics when developing high performance computer systems is just as the proverbial engineer that said he is busy designing bridges, and has no time to learn about rivets, welding and bolts.
Sorry Josef, but I just now saw this interesting question...
The underestimation of basic electronics in IT specialties is my main problem because I teach exactly such IT students. I solve it at the cost of much efforts by trying to convince them that this enters into the common technical literacy of Engineers what they will become. To motivate them and draw their attention, I often say something like, "It exists in every modern computer..." But in general the work of teachers with these students is a thankless job...
According to me, these students should be familiar with basic concepts in circuitry without deep penetration in details. Especially important are the peripheral electronic circuits - different types of transducers, amplifiers, sensors, measuring circuits...
"But in general the work of teachers with these students is a thankless job..."
Yes, Cyril, IT students think that in the real world is enough keyboard for "all problems". But the "reality" is very hard, and keyboard alone is "hardly enough" :))
Many current 'hot topics' in IT (eg. 'big data', 'swarm intelligence' - just to name two of them) do not require an understanding of the underlying circuits.
On the ither hand, embedded control (a vast field) is hardly covered satisfactorily without - more than 'basic' - knowledge about the electronics involved.
This is the experience from quite a number of years in 'systems development' - embedded as well as systems where the hardware doesn't matter.
Thanks for invitation. A good policy based question.
The student do deal with creativity need to understand both very well at deeper levels. the student do deal with normal programming need to understand both at moderate level. The student do deal as user need to understand both at basic level.
Regardless of the academic dispute of splitting the discipline into computer science and engineering - a valid but moot point- once one enters the field of applied informatics such as process control, data acquisition and programmable controllers of any kind, I believe electronics remains a fundamental pylon . In this very physical layer of industrial, immediate application all abstraction programmers are accustomed to evaporates to leave behind direct hardware related calls, and that leads to the need for a good understanding of the basics.System engineers need to know that perfect system they just designed has sensor-related tolerances and non-linearity, needs compensation of different types and degrees and linearizing circuits have their limits, that look-up data tables just make for an easier implementation of a very computing-intensive equation on low power distributed controllers, etc.Even when considering a new digital design for a new type of measuring device, using some sort of sensor, questions of how fast the system needs be, what bandwidth etc- find relevant answers in basic electronics not in digital design. While it is simple to maximize everything sometimes low-cost, low part number designs fully cover the possibilities of the sensor- and help reduce cost, and unnecessary overhead in more complex systems they will be integrated into.
My opinion is that someone neglecting basic electronics when developing high performance computer systems is just as the proverbial engineer that said he is busy designing bridges, and has no time to learn about rivets, welding and bolts.