Archive for October 23, 2020

Computing “Revolution” of Past Two Decades as a Showy Failure: 1

October 23, 2020 40 comments

One of the defining features of the past two decades in west has been the dominant position in public consciousness of corporations involved in manufacturing personal computer hardware (desktops, laptops, smartphones, tablets, IoT crap, embedded electronics etc) or making them function and do stuff (‘IT’ corporations such as Google, FakeBook, MicroShaft etc). One could say that Amazon is a an ‘IT’ company which sells stuff that people used to buy in department stores. A large part of current market value of many stock indices in the west now comes from corporations who either make personal computational hardware or the software they run.

But have you ever asked yourself- has these rise of these corporations or the widespread usage of products and services sold by them actually improved the quality of life for the vast majority of people. To understand what I talking about, let us ask two more basic questions. Question #1: Would the absence of personal computing “revolution” during past twenty years have any negative effect on the quality of life or somehow constrain development of other technologies? Question #2: Has the computing “revolution” improved quality or reliability of other products and services, let alone increase the general quality of life for vast majority? As you will soon see, the answers to both questions are obvious as well as surprising.

The unpleasant fact for many geeks is that the computational ‘revolution’ of past two decades has been the most sterile and unproductive period of general technological advancement in the past two hundred years- and I do not make that claim lightly. To better understand what I am getting at, ask yourself if you can name a single non-computer product that has improved your life or is somehow associated with the modern world which would not have existed without this pseudo “revolution”. Give it a try.. can you think of any non-compter product which would not have exsited without this so-called “revolution”.

Since we have to start from somewhere- let us start with modern jet airliners? Well.. every airliner designed until the late 1990s was largely designed by competent engineers using their engineers using their experience and some combination of slide rules, desktop calculators, 8- or 16- bit desktops connected to a few clunky mainframes. The DC-9, DC-10, 737, 747 etc were designed in what was essentially pre-computer era. The A-320 was designed at very start of era where electronic computers (mostly mainframes) of any type were widely used for aircraft design. The 777 was the last aircraft designed with a combination of good engineering and primitive CAD technology. Only the 787 was designed in era of modern “computing”- and it has been the most over-budget and troubled design of them all.

And this is not just restricted to airliners. Consider space exploration and missiles. The space race between erstwhile USSR and USA occurred before the modern computing “revolution”. People went into space before even their vehicles had a single solid-state transistors, let alone a IC or CPU, within their rockets and vehicles. The flight control computer used in Apollo missions was a hand-made computer with about the same computational capability as an early Apple II, TRS-80 or Commodore PET- though it was a 16-bit machine. The Pioneer and Voyageur Probes which are the only man-made objects to visit Uranus and Neptune (albeit in a fly-by) did not have CCD cameras nor CPUs. The same is true for both Viking probes which landed on Mars in l970s as well as the Venera family of space probes that USSR successfully landed on Venus in that era. Oh.. and all those lunar probes and soviet lunar rovers too.

The vast majority of space probes launched prior to late 1990s used tube technology (or very primitive CCDs) for imaging and very basic IC circuits joined to make ersatz CPUs. And guess what.. they performed their job magnificently. But it gets even more interesting when you look at aircraft and missiles used by the military. Did you know that first ICBMs did not use solid-state electronics and it was not until the 1980s that ICBMs using Integrated Circuit Blocks for guidance became commonplace. Funny thing is that the accuracy of ICBMs has not increased by a worthwhile margin since the 1980s. Even ALCMs (Air Launched Cruise Missiles) achieved almost the same accuracy and guidance capabilities as those used today with what essentially a mixture of custom ASICS along with 8- and 16- bit CPUs. The GPS system worked just fine with receivers that contained what were essentially 8- and 16- bit CPUs.

Even the state of design for nuclear weapons, which were often designed using a combination of previous experience and calculations on some of the first real “supercomputers”, has not progressed much further than it was in the mid-1980s. Remember that every single warhead in American and Russian Inventory was (at best) designed on a “super-computer” with less computational power than the original XboX game console. The same holds for design of everything from nuclear submarines, tanks, guns and missiles. To put it bluntly, even in areas where the computational “revolution” should have helped the most, things have been pretty stagnant since the 1980s- and not for the lack of money and resources thrown at the Military-Industrial complex. It is as if big and substantial technological advances haven’t occurred in these and many other fields since the late 1980s to mid-1990s.

Since we are at almost a thousand words, I will wrap up this post. In the next ones, I will write about how the so-called computational “revolution” has not improved the quality of housing and automobiles, school and university education, transport and corporate logistics, process of drug development, everyday financial transactions and.. yes.. even popular entertainment. Even popular entertainment..

What do you think? Comments?