Comments by "Keit Hammleter" (@keithammleter3824) on "National Semiconductor: "Animals of Silicon Valley"" video.

  1. In the 1970's, after a career in electronics repair, I became a new electronics product design engineer. An engineer needs data books and application guides on chips. The major chip companies then were Motorola, Texas, Fairchild, and National Semiconductor. Texas delivery times were way too long. Philips was trying to nibble away at the edges. So I asked the local reps for Motorola, Fairchild, and Nat Semi for data books and application guides. The Motorola guy refused to give me anything, saying he had supplied the company Head Office library. Fat lot of good that is - an engineer needs everything to hand at all times. The Fairchild guy gave me a TTL data book and a TTL app guide. The Nat Semi guy gave me a truck load of data books and manuals. Extensive literature on each chip showing how to use it and get the beast out of it. And quoted lower prices and faster delivery than Fairchild. Guess which brand of chips got used the most in the products I designed .....
    28
  2. Nat Semi's analogue chips were great because they were designed by Bob Dobkin, Bob Widlar, and Bob Pease - all clear thinking brilliant men. When Nat Semi turned to things like calculators and microprocessors, these guys were not involved, and the resulting products were junk. For example their calculators were Reverse Polish stack entry, same as the earlier brilliant and very successful Hewlett-Packard calculators. But HP had done a study which showed that a stack level of 4 covered just about any calculation you would want to do. But Nat Semi calculators had only a three-level stack, which made them useless. Another example: Nat Semi brought out a microprocessor, the NSC800, that was in hardware a CMOS Intel 8085 but with a Zilog Z80 instruction set. Brilliant idea - low power, cheaper printed circuit board needed, and the Z80 instruction set was far more powerful than Intel's. But Nat Semi omitted the 8085's serial data ports. Dumb. It meant no-one wanted it. And the main advantage of the 8085 came with using the combined memory/port chips, which Nat Semi couldn't supply. And there was no second source, so nobody risked the NSC800.
    2
  3. Asianometry implied the 16000 series was Nat Semi's first microprocessor. This is not so. the 16000 series came out after a long line of microprocessors from Nat Semi, beginning with the PACEr, weird toys eg SC/MP, and second sourcing Intel micros such as the 8080. The 16000 series was a brilliant design but failed to sell because a) it came out after Intel and Zilog had captured the market, b) there was no second source - its a big business risk if you base your products on parts you can only get from one supplier - you could, if Intel's factory burnt down or something, get the same thing made under license by NEC or Philips; c) the 16000 chips were full of major bugs.
    2