The Straight HP 9825 Story, from Don Morris
“Bob Watson, Tom Haswell, and I were always worried that we just might reach a river over which we could not build a bridge. IC design came to a halt because there were no design rules or performance specs. Why? Because there were no working test transistors. Even if there were a process to build the transistors, we would still fail because no one was supplying IC masks of adequate quality. Even if the mask vendors got their act together, we would still fail because the defects in the glasses (for the masks) available at the time were too large for our uses.
I got so depressed that I could hardly keep the secret of our imminent failure away from the troops any longer. So Fred Wenninger sent me to talk to Watson. He said that Watson was the expert at skating on thin ice and knowing which bets to make. He also said that Watson was still putting new money on the table.
Watson told me about the bipolar chips in the 9800 processor that didn't come through until just days before the project would have been canceled. So I explained that to Watson that for him to bet and win once wasn't adequate reason to expect it twice.
Then Bob did something amazing as a manager. He got me reinvolved 100%. How? He told me that my crew and I were likely the luckiest engineers in the world, or at least at HP, because we could work, be paid, and ranked based on our own descriptions of our results. Management would have to take our word that we were building the perfect machine if it couldn't be shown otherwise because the ICs couldn’t be fabricated. He explained that my career’s continuation would be based on the execution of my responsibilities and Tom's career on the execution of his responsibilities. Watson's, career was riding on whether he could pick the right people to trust.
I went back to my desk and called Dave Maitland over. I decided that if there was never an IC process, then any design rule would be adequate so long as we could say that the designs were finished. So I asked Dave about design rules ... knowing full well that the original NMOS I ROM was designed by Dave prior to having NMOS I design rules. He explained that the BPC and other chips we were developing couldn't be designed with the cavalier methods he used previously. Real NMOS II design rules were required. But Maitland allowed that he was smart enough to deduce what the rules would or should be. We spent about an hour at my desk generating the NMOS II design rules that we handed out to the IC design engineers.
Then the physical IC designs really got underway! The NMOS II meetings that followed were something else. Usually Fred Wenninger attended them for me* so that I could more easily stonewall. I said give me real design rules and we will use them, but until then I will use my own. From that opportunity flowed forth literary masterpieces from Tom Haswell that fell somewhere between the Bible and Shakespeare. In these letters, he described the proper ways to develop production IC processes and pointedly wrote that I had circumvented all these ways. If the NMOS II ICs failed to work, he wrote, it would be all my fault, not his! Haswell’s literary works were to cover his ass under all circumstances. But Haswell and I got along anyway, and met together often. While our wives would take the kids away from these disruptive discussions, we would work on our opinions of the other's professional performances. Oh the days!
As it turned out, I was right about the design rules. I told Tom that I was saving him gobs of time and effort because he normally would be looking for optimal configurations, but now he now only had to run one set of test mask designs and verify that he achieved the planned performance figures. That turned out reasonably simple because the process turned out to have a lot more performance than the original NMOS II IC designs required.
Bottom line: Bob Watson's career was salvaged for knowing who to pick, or was it being lucky a second time? Not for me to judge. But as I read the histories on this Web site, it looks like everything was logical, expected, and that the whole crew knew from the beginning that we had a winner. We knew if everything worked, the HP 9825 would be a winner, but there were several of us who knew it was never truly in the bag. We put it in the bag, with great difficulty.