For novel ideas about building embedded systems (both hardware and firmware), join the 40,000+ engineers who subscribe to The Embedded Muse, a free biweekly newsletter. The Muse has no hype and no vendor PR. Click here to subscribe.
By Jack Ganssle
Published in Embedded Systems Programming, August 2003
The arthritis has nearly incapacitated me; gout and advancing Alzheimer's number my remaining days. But one of the grandchildren drives me to the cemetery so I can make my yearly pilgrimage to the tombs of old friends. Leaning heavily on the cane, I clop my weary three-footed shuffle through the well-tended grass that disguises the sad rot below. The present age is a ghostly mirage to me now; the past seems more alive, many long-departed friends visit in my sleeping and waking dreams.
I make my way to Joe's government-issue headstone. Slow tears come unbidden as I knew they would, as they always do. The pain makes me wince as I lean over and place flowers on the damp earth.
He lies forgotten in this potter's field, having died penniless and without caring family. Like so many other highly-educated people, his career, hopes and dreams all crashed during the devastating second decade of the 21st century. I alone had the wisdom to invest in China and Bangladesh during the early days before engineering all but disappeared from this country. Great wealth, though, hasn't tempered the bitterness as I recall the topsy-turvy upending of the world as we Westerners understood it.
America now lies in ruins. The cities have all been abandoned, what remains of their streets blocked with the rusting carcasses of a once mighty industrialization. The skyscrapers which once vaulted heavenwards in an earlier age of hope are now ruins in an age of despair. Scavengers compete with the rats for a bare subsistence, though the Federal government long ago declared all cities no-trespassing zones.
Everyone lives in rural areas. This vast land, once crisscrossed with a million miles of gleaming highways, supports a network of family owned and run small farms. We're again an agrarian society. Trade flourishes despite the IRS's attempt to tax barter. An almost non-existent tax base has shrunk federal employment from its high of over 2 million to a few thousand today.
Joe - poor Joe! - was one of the millions of victims of the collapse of the industrial revolution. We graduated from college together, our engineering degrees then a guarantor of success in the middle-America dream of modest prosperity, security and relative ease. But neither of us became engineers for the money; we both were drawn to the profession by a love of gadgets and a desire to build things. It seemed almost a scam that companies would pay us to design stuff. But pay us they did. Before the first big crash in the late 90s (remember "dot coms"?) our high salaries augmented by a lose California lifestyle made us the envy of our liberal-arts friends.
If only we engineers had had a sense of history - even recent history! I can recall when this country was a manufacturing powerhouse, when our plants and factories were the envy of the world. In my youth "Made in Japan" meant cheap junk. Secure in our mechanized dominance, like Romans of old we (as we once said) "dissed" the rest of the world's capabilities. Our parents, who fought the second of the three world wars, always faced Asia with a subtle racism which translated into a feeling that "those" people could never seriously compete with us.
Yet compete they did. In the 70s Japan reinvented manufacturing, cranking out cars of superior quality for less than their American counterparts. In a few short years imports devastated our auto industry.
Joe and I were kids then. We studied the books cranked out by armies of management consultants returned from their studies of the Japanese carmakers, and like so many others thought we had learned the lessons needed to compete successfully in this new world. Quality was job one. Invest in tooling rather than blue-collar workers. The short-term result gave us awesome vehicles offering features previously unimagined, which ran for years with minimal servicing. Pundits promised that those autoworkers displaced by the robots could be retrained to create web pages. Some even believed it.
Many of those people never returned to a skilled profession. How was it we missed the message of legions of hamburger-flipping Bachelors of Arts folks? Though jobs abounded most offered below-sustenance wages. For generations immigrants to this country pushed their kids to get a better education than they had had. But that schooling, unless in one of a few highly specific technical arenas, no longer correlated to success.
Joe and I were undaunted. Our hearts bled a bit for the displaced. We complained about the increasing taxes going to the unemployed. But could design great electronic products. Many of our colleagues lost their jobs forever in The Great Recession of 2000, but most of us survived. The workload increased, schedules became even more unrealistic. Consumers demanded ever more from us; we responded with smaller, better, cheaper products.
Around 2005 the ASIC industry failed. Some say that event, nearly unnoticed at the time, was the crack in the dam that precipitated the loss of electronics to the (then) third world. Why were we so focused on Moore's Law? It had become an end in itself. Semiconductor vendors pushed their processes from 150 nanometers to 130, then to 90, with sights set on 45. Billions went into new fabs; mask sets cost millions. Transistor counts soared to the hundreds of millions on a single chip.
Though the technology was awesome consumers just did not care. With costs so high only a very few applications could benefit from the astonishing level of integration. Venture Capital dried up since the VCs only funded outfits that could service those few apps. No new capital meant no new jobs, which translated into a continued stock market slump. Belt tightening turned into corporate bulimia. The US economy staggered along, shedding jobs. The long-awaited recovery never happened. Consumer spending started to dry up.
CIA reports revealed that an old chemical weapons factory in Pakistan had been retooled to produce chips using surplused steppers. That was considered a Good Thing at the time. Its 200-nanometer technology was the laughingstock of the semiconductor industry. But the company modeled itself on the Russian's brute force - but reliable and cheap - space program. Iran and the former Iraq followed suit. A $6 CPU that ran at 30% of the speed of the latest $500 whiz-bang device proved itself good enough for most of the market.
Intel and AMD dropped off the big board as the PC industry surged overseas.
All of the semi vendors tanked, their fabs now white elephants. CEOs of fabless chip companies were briefly business superstars as they congratulated themselves on avoiding such huge investments. But as wafer plants were shuttered across the US, and then in Taiwan, Singapore and Mexico, they found themselves unable to produce their products.
By 2010 Middle East semiconductor sales outpaced their oil revenues. The University of Islamabad replaced Cal Tech and MIT as the center of engineering learning. More US engineers lost their jobs.
Yet some of us thrived. We still had the software industry. Only Americans could produce killer code. I made my first overseas investments. Joe worked harder, putting in long hours due to corporate necessities and sheer love of writing firmware.
Some say H1-B visas presaged the hemorrhage of software jobs. I'm not so sure. The law was written to ensure that visa holders would get competitive salaries, so their skills would augment the market rather than undermine employment of citizens. But as living conditions became less tenable here due to the vast number of displaced workers in other industries, and as the unions made life difficult for anyone not holding a Federal Citizen Identity Card, a flood of former H1-B holders left, returning to their homelands, and taking the skills they learned here with them. Historians claim a vast wave of intellectual property flowed overseas at the same time.
Ramesh Indira's invention of the DWIMNWIW (Do What I Mean, Not What I Wrote) compiler drew on his work at Sun and Mentor. Coupled with India's long-established tradition of using a well-managed software process, and the secrecy surrounding his compiler, an explosion of cheap yet innovative products flowed from Bangalore. The long-established middle-class profited; a surge of consumerism from that nation's poorer people propelled India's economy to the world's largest.
"American software" remained a convoluted mess, and became an epithet equivalent to the old "Made in Japan". Microsoft was the only glue that held the US software industry in any semblance of financial order. Long struggles with Linux and other open source competition had resulted in an OSS version of Windows. Microsoft, too, succumbed when an enterprising 14 year old in New Delhi ran Windows through the DWIMNWIW compiler, removing all of the buffer overflow problems that had plagued the product for so long. Political and ethical bodies had never had the courage or wisdom to address the rampant piracy enabled by the net, so the perfected OS had over a billion downloads in its first day on-line.
Joe's superprogrammer status helped him survive the layoffs. He worked at a small company which squeaked by only via long hours and wave after wave of salary cuts. His marriage tottered and then collapsed as he buried himself in his work. Estranged from his daughter, his increasing bitterness with his personal life spilled over into work. Abrupt with his peers, cynical with the boss, haggard and overworked, he, like so many others with minimal health care, succumbed to the influenza epidemic of 2018.
As I foresaw, even India's success was doomed. A hundred years of the Raj had created a yearning for more goodies, more success, more wealth. Salaries escalated on a par with increasing success. We humans are genetically pretty much the same, regardless of where we were born. Smart people abound; instant communications removes all regional barriers to business. Jobs inevitably migrate to the lowest cost providers. China's dominance of the software industry elevated their population to a brief moment of affluence. Today Bangladesh and equatorial Africa are the world leaders in software. Tomorrow? Who knows?
I'm a withered old man of 58 now. Few survive so long in these days. My mandatory euthanasia call arrived from AARP this morning, so I'll soon join Joe in his slumber. I'm ready. I loved my engineering career, and was lucky to have practiced it in its heyday. Joe and I lived in a Camelot-like bubble of prosperity and ease. My heart bleeds for young people today, those smart folks who will spend their weary days scratching a living from the unforgiving soil.
There's wealth for engineers in Africa today, but the signs of collapse are appearing. Increasing wages mean companies, in their relentless hunt for profits and low-cost suppliers, will start looking for other options.
I'm telling my grandchildren to invest in Franz Josef Land.