Technology - see the future from history

[tw]

As I begin to write this, I suspect this will become so long as to be multiple posts. We learn from history. This will demonstrate how to learn even how to invest by addressing the only thing that makes jobs, industries, markets, products, wealth, and happiness.

In designing semiconductors for a calculator, Intel stumbled upon a concept nobody had considered. A microprocessor. Intel bought the design back from that Japanese calculator manufacturer (who soon went bankrupt). And developed products with famous numbers. 4004. 4040. 8008. And the most famous - 8080. The 8080 became the basis of uprocessors that would be on most every home. Intel had only one problem. Wall Street was becoming dominated by business school types. Stock brokers were incapable of investing in innovation due to their business school training.

In the meantime, Intel was doing what any patriot company does. They developed semiconductor memory. Static RAMS (ie 2104). Dynamic RAMS (ie 1103). Non-volatile memory (2704). Single chip computer (8051). Single chip peripherals (8251). and non-volatile programmable single chip computer (8751). In each case, first two digits defined the design family. Notice a difference between a uprocessor and a single chip computer - a significance that will have relevance later. Most parts found standard in a computer were originally pioneered by Intel. Intel also attempted to break into other businesses (motherboard, RealTime Operating Systems, household appliance networking, automobile electronics (ie CAN bus), multiprocessor computing (Multibus), etc. Many of these businesses had no legs. Intel sold off many (memory, peripheral, single chip computers) to concentrate on what it did best.

Intel also pioneered new business models such as DIX. The consortium of Digital Equipment (ie VAX), Intel, and Xerox (Apple's MacIntosh) that created a standard we still use today. Ethernet. An Intel created standards so that other companies could contribute what they did best (ATX standard).

Fundamental to what Intel did was something Andy Moore created as an afterthought for an article in Electronics Magazine. Moore's law became a benchmark to identify business school myopia before it adversely affected the spread sheet. Moore's Law went a long way in saving Intel. Meanwhile those other two DIX partners feel victim to what is taught in business schools. DEC stopped innovating with the VAX. Xerox had begun to stifle innovation when the DIX standard was created. By the time business school thinking was identified, the damage was so great as to even affect the spread sheets.

Intel's history is full of upstart competitors who sometimes had superior products. An early one was Zilog (Z-80) that was superior to the 8080. Intel's answer (8085) was inferior to the Z-80. But Intel also innovated elsewhere. Intel abandoned the 8 bit microprocessor market by making an 8080 equivalent in a 16 bit uprocessor (8086). That processor was recommended by Bill Gates for the original IBM-PC. And that processor finally meant Intel could get investment money to build a Fab. To make processors on an industrial scale. The Silicon Valley did not yet develop another new business model - venture capitalists (Sand Hill Road). Wall Street was already so business school brainwashed as to not invest in Intel. IBM's investment to make possible an 8086 production line is why Intel was finally able to become profitable.

Keeping Intel productive were other challengers. Zilog was a first that could not innovate beyond the Z-80. And was eventually bought, and then dismantled, by Exxon. The next was Motorola's 68000 line. Anyone who thought they knew computers because they were programmers love the 68000. All its (16?) registers could do counting, access peripherals, and memory indexing. Anyone who actually knew computer immediately recognized the 68000 as a disaster for the same reason. An example of two completely different conclusions from the same fact. But only one is accurate due to also learning background and history. Apple abandoned their underpowered Motorola 6805 for the 68000. Even software had to be completely rewritten. Eventually, the defect in Motorola's 68050 was so obvious that Motorola canceled that processor after completing the design.

Bottlenecks in processor are speed (processor speed, memory size, and memory access), power (heat), and software. Zilog made a processor that ran all Intel 8080 software, was faster, and consumed less power. Motorola threw lots of money keeping their slower architecture as fast as Intel's, and had to develop all new hardware (including peripheral chips) and software from scratch. Motorola just could not complete against a company that kept benchmarking itself against Moore's law.

The POWER had been developed in Austin TX as the next wave of innovation. About once per decade, IBM would scrap their entire processor design to build something completely new. The 360 Series being the most famous. But even that was replaced by the 370 series. Then business school graduates took over. What should have been the next architecture in the IBM line was the POWER series. But business school graduates routinely stifle innovation in the name of technical ignorance and fear.

Apple, also in the process of self destructing, ran to IBM for another processor. A stripped down POWER processor renamed the POWER-PC. Apple had even built and demonstrated Macs using Intel's processors. But, like at IBM, business school graduates could not understand the advantage; only saw fear. We all know how bad Apple was becoming, in part, because it used an inferior processor. Eventually, IBM could not compete with Intel. Intel had a benchmark based in the product. IBM was using spread sheets and business school concepts to first destroy the IBM-PC business (Entry Systems Division). Stifle innovation (VGA monitor, OS2, Micro-Channel). And attempt to make profits at the expense of all other industry partners.

While Motorola was failing, IBM was also so pathetic that in 1990, none of the PCs in corporate could execute any retail software. IBM was still using 1984 hardware in 1990 offices - at least two generations obsolete. Their president and the head of their PC division did not even know how computers work. IBM would blame Intel, Microsoft, PC clones, and even the Japanese for their problems.
{continues in the next few posts}