Part of the Silicon Valley died this week when National Semiconductor got bought by Texas Instruments. It was another box checked off on the long slow autopsy of the golden era of semiconductors, an era begun in the 1950s when companies starting supplanting fruit orchards in the Santa Clara Valley; when National fled the East Coast for Santa Clara to be where the action was.
For generations, National invented and grew, and the communities around it benefited when National paid taxes that helped build roads, schools and sewer systems, employed thousands who built communities and nurtured schools and spawned and supported local businesses. Nearby a similar story: Intel, AMD, Linear, and so on down the storied honor roll of Silicon Valley pioneers.
Even as National seemed to peak in the 1990s; even as it wobbled through the Gil Amelio regime (what the heck was that all about??) and tried going toe to toe with Intel (Brian Halla eventually saw the light on the strategic error), the company and its people were always a cornerstone of the Valley.
And they had fun. Bernie Cole writes about faux charges up San Juan Hill; Paul Rako has a fantastic look back at National and Valley history, through the eyes of the late Bob Widlar (two of the outstanding photos in that post tell you exactly what Widlar would think of the merger).
It’s all washing away now in a warm shower of billions of dollars from the Texans. There will be optimistic talk in the coming months about synergy and strengths and mutual dependencies, of scalability and leverage and every other vacuous, meaningless messaging mantra you can conceive of.
The words will fade as hundreds of jobs fall, like cherry blossoms, and 1152 Kifer Road becomes a satellite office rather than a power center. Don’t be surprised in five years if you drive past it and it looks just like Motorola SPS’s old power center on North 56th St. in Phoenix—today, a gleaming… white… and completely empty reminder of a vanished time. There, too, the operations moved south to Texas.
Back to the Valley. Last year, MicroSemi, an aggressively expanding analog house, snapped up Actel, one of the Valley’s FPGA pioneers. Lots of talk about minimal redundancies and so on, but scores of people have lost their jobs and those remaining are being pressured to move south, to Southern California, to the headquarters of the winning team, to the… Not-Silicon Valley.
So, we mourn the dead.
And then we move on.
We move on because, in the distant warm memories of a fading few, families packed picnics and piled into old trucks and station wagons and rattled up Mt. Hamilton or into the Santa Cruz Mountains at springtime to marvel at the impossibly vast vista of blossoming fruit trees carpeting the Valley below.
Replaced by vast tangled freeways and thousands of bland buildings stamped out by armies of architects and builders who, long before, abandoned the notion of excelling at their craft.
We move on because what replaced the farms was an economy of unimaginable innovation, wealth, optimism and embrace of creative destruction… an economy of National, Intel, AMD and, later, Xilinx, Altera, Actel, Linear and scores more.
We move on because of the same creative destruction that companies like National feasted on in their youth; we move on because the Valley will always be the cradle of innovation. After the silicon guys came the systems guys; now, the software guys—Google. Facebook. Twitter. They all stand on the shoulders of National Semiconductor and other Silicon Valley pioneers.
To: "brian fuller"
Cc: "junko yoshida" Sent: Wednesday, May 11, 2011 8:50:21 PM
Subject: requiem for an era
Hello Mr Fuller,
I spent most of WWII in the SF area.
Fort Berry, the Presidio and Berika Street.
My mother and I visited the carrier Franklin on its return to SF.
I was a student at Commodore Sloat school in SF for about first and second grade.
We moved to San Mateo in about 1947.
I was a student at Beresford grade school in south san mateo.
Diane Varsi was a classmate.
Then Borel middle school. I was there for the days of '49.
I attended San Mateo high school.
Enviromental damage done by silicon valley and the freeway system was great.
Electronic Engineering Times, January 22, 1996, p. 84.
AND since I am MAD about this I would add, try to find a simple part now like a 3046 transistor array.......??
The industry is now MARKET driven which means if the Chinese robots are not chomping about a million of a line item a month the fat cats in TI marketing will probably delete it......!!!
National's databooks and appnotes were always excellent. For authoritative comment see all input on web from Bob Pease and Paul Rako, both ex-NSC employees. The best linear appnotes in the business are by Bob Pease and Jim Williams (LT).
TI will just delete useful NSC parts and the creative electronics industry will be the WORSE for it. DAMN !!!!
I've ALWAYS had problems trying to find data on any product made by a company that was later absorbed by another company. Almost certainly the same will happen to NS. I'll be hanging on to my 1980 data books....
RWatkins, you make a fantastic point. Something like 75 % of all mergers fail. (And I'd love to see data that slices the results by size of merger)...
As journalists, we should probably start every merger story by writing "Three quarters of all mergers fail. Why is this any different?"
All of these comments and no mention of what most of us "old timers" are worried about. We hope that TI purchased National to get a better grip on the market, not to eliminate its competitor. We have seen TI purchase other entities and create nightmares for designers. Two good examples are TI's Burr-Brown and Luminary purchases. In each case, products were assimilated poorly and details were lost in release of new products that drove designers away from TI when stuff that had historically been high quality did not work, did not perform per data sheet, or worse experienced early failure due to chip design issues not resolved prior to release. I personally felt this pain designing in a Burr Brown ADC (first problem EVER using Burr Brown product) and a Luminary processor (separate projects). Let us all pray to the TI bean-counter gods that they will not lay off critical engineering staff that had kept National product releases and quality exceptional, and they will not force premature release of products that were in the pipeline at time of purchase.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.