What are the ten best decisions ever made in the semiconductor industry's history? This is the stuff of which long evenings are made, but here, for what it's worth, is one list compiled after a long evening.
- Bell Labs' decision to sell licenses for transistor manufacturing technology
- Akio Morita's decision to buy a transistor manufacturing license from Bell Labs.
- William Shockley's decision to hire Bob Noyce and Gordon Moore
- Fairchild's decision to make ICs.
- The decision to invest $3m in founding Intel
- Intel's decision to make silicon gate MOS memory.
- Taiwan's purchase of CMOS technology from RCA in 1976.
- Chips and Technology's decision to clone IBM PC chip-sets.
- The decisions by Altera and Xilinx to found the fabless semiconductor industry business model.
- Philips' decision to co-found TSMC with the Taiwan government and kick-start the foundry industry.
Next week: The Ten Worst Decisions Ever Made in the Semiconductor Industry.
A couple of quibbles (on a short evening...):-
Getting Fairchild and TI (an aerial camera company and a geophysical intrument company respectively) into the semiconductor business has to rate up there. TI is still a big factor.
And the fabless semiconductor company model existed in 1972, before Altera or Xilinx were thought of. I worked for one such company that year, and I am pretty sure there were others.
Also, surely the decision to build a complete computer core as an IC was big. Yes, I remember looking at a teletype in the middle 1960s, and thinking that it could be turned into a much smaller computer than the ones I was already using, with ICs, but actualy doing it (the 4004) was significant.
How about the 15 best decisions?
OK, if we're going for 15, then let's add:
The decision to build a complete computer core as an IC (the 4004)
The 1976 decision by Japan to embark on the VLSI Project.
The 1987 decision by the US to embark on Sematech.
The 1983 decision by Philips and Siemens to set up the Megaproject (which morphed into Jessi and Medea).
The 1970s decision by Monolithic Memories to make bipolar PALs which kick-started the programmable logic industry.
To what degree are those "decisions" - or "best" ?
A lot of them are "things that happened" - but would the world be that different if they hadn't?
Indeed, perhaps things could have turned out better...
#1 is definitely a decision, and if AT&T had not licensed (which they easily could have done) then the world would be very different.
Some of the others were good decisions for the company that made them (eg Intel), but others seem a bit ho-hum.
But were founding VLSI Project / Sematech / Megaproject really that significant?
C&T doing clone chips was good for them for a while, but not industry-changing someone else would have done it anyway. Ditto MM making PALs.
I would put two different ones on the list:
The decision by IBM to license the PC. Without that we would have no PC-clone industry, less economy of scale and smaller volumes of more expensive home computers.
The legal decisions that opened up the phone network to customer equipment - arguably the first & most significant "democratisation". Without that, we would still be restricted to operator provided handsets, and the provision of fax, modem & broadband (with all the semiconductor volumes that resulted) would all have been far-far slower.
To me the AT&T license decision, the VLSI project, the Megaproject, and Sematech all advanced, accelerated and broadened the industry, technically and geographically.
Probably everything that has hapened since would have happened anyway but not, in my view, so quickly.
On the Chips & Technologies point our recollections of industry history obviously differ because I didn't think that IBM had licensed the PC.
IBM had used a 3rd party O/S, MS-DOS, and a 3rd part microprocessor, the Intel 8088, which meant that the PC was clonable with a cloned chip-set (plus a reverse-engineered BIOS which Phoenix Technologies provided).
Which is why, with Chps & Technologies providing the cloned chip-set, I thought they had made such an important contribution because they enabled the IBM clone industry.
To my recollection it all happened without IBM wanting it to happen.
You are absolutely right about the legal decsions which opened up the telecoms industry with huge benefits for the chip and equipment industries. I was containing my Top Ten to purely semiconductor industry decisions.