Looking Back on My Engineering Career, Part 1: Two Large Companies
A brief* professional biography
I’m on the verge of retirement (less than two years) so it’s a good time for me to look back on my career.
Ack, what a boring topic! But it’s content. Maybe some of you fellow old codgers will enjoy a trip down memory lane. Maybe some of you young whippersnappers will learn some industry history. I’ll try to make it a little bit amusing.
Besides being boring, this history got very long, so I’ve broken it into three parts. Here is Part 1, about when I worked for Two Large Companies.
My entire career has been spent in computer hardware and software engineering. My trajectory has been maybe a little different than most, with a mixture of a bunch of different stuff. But then, not so different at all, since it is all computer engineering. It’s not like I’ve ever done construction or farming or anything really different, which I somewhat regret.
Let’s start in high school, where a physics teacher (LOL, high school “physics” indeed!) predicted I would become a physicist. I graduated in 1973, just to anchor the timeline.
Then off to college at the California Institute of Technology (Cal Tech), the brainiac university in Pasadena, California, just a few miles from the San Fernando Valley where I grew up. (Los Angeles area, for those who aren’t familiar.) I had no idea of how tough Cal Tech would be. I instantly fell from being the smartest to the dumbest student in my class.
They say 90% of freshman at Cal Tech are physics majors, and 90% of seniors are engineering majors. That was my story. When I hit calculus with more than three dimensions (it goes to “n” dimensions after that), my head exploded so physics was off the table.
After a couple of years, my best friend from High School convinced me to transfer to U.C. Berkeley with him. He had gone to Pepperdine Malibu because they had the prettiest girls. He was right about that, but none of them would give him the time of day! I joined him in Berkeley to get out of the brutal curriculum at Cal Tech and also to broaden my horizons beyond Southern California.
When I transferred schools and signed up for some engineering classes, I was told, “You can’t take that, calculus is a prerequisite.” I had to convince them that Math 1A at Cal Tech IS calculus! People seem to think that Berkeley is a tough engineering school, but I was shocked how easy it was after Cal Tech.
College was a drag. The engineering workload was heavy and I was not a very socially successful person. I thought about dropping out. After my Junior year, I took a six month paid intern position at IBM in San Jose. I tested transformers in an incoming inspection lab. It was so boring, it motivated me to finish my degree.
I won the Smokey the Bear award one time for blowing up an oscilloscope. I forgot to put the high-voltage attachment on the probe before testing a high-voltage pulse transformer. I was terrified about the oscilloscope but they just laughed and tagged it for recalibration. LOL
My first job with my engineering degree was at Intel on a microprocessor design team. I moved into Intel’s brand new Walsh building across Central Expressway from its headquarters on Bowers Ave. in Santa Clara, California, a.k.a. Silicon Valley. It was Intel’s second ever office building after the headquarters building. The HQ building includes a small fab line (I assume it is still there) in addition to marketing and engineering offices. The Walsh building is strictly a cubicle farm, very few hard-walled offices. Cubicles were just arriving on the scene at the time.
Intel had a strict 8 to 5 mandatory working hours policy (which did not, of course, prohibit unpaid overtime hours). They would periodically have a sign-in sheet at the security desk that you had to sign if you came in after 8. Common names on the sign-in list were Mickey Mouse, Goofy, and any of the company founders (Gordon Moore, Bob Noyce, and Andy Grove; Gordon and Any could be seen frequently around the offices, but not so much Bob).
I worked on the 8021 and 8049 microcontrollers and the “RUPI”:
- The 8021 was a little brother of the very popular 8048 microcontroller. It sold for 25 cents in volume and went into applications like microwave ovens.
- The 8049 was a big brother to the 8048, doubling the amount of on-chip RAM and ROM. We fucked up and expanded the die in the wrong direction so it had to turn 90 degrees to go into the package, thus preventing it from being pinout compatible with the 8048.
- RUPI is an acronym for Remote Universal Peripheral Interface. Basically I designed an HDLC/SDLC communications controller and stuck it onto the side of an 8048 microcontroller to support networking.
This was in the days before Electronic Design Automation (EDA) systems. A bunch of women (it was always women for some reason, mostly South and East Asian), laid out the masks for the chips manually using colored pencils on large sheets of mylar. I did a lot of planning and checking of the layouts. One middle-aged Indian lady doing layout work got a hoot out of sexually harassing the young male engineers. Nothing serious, just joking around to embarrass them. Good times!
Intel was growing like crazy in those years and they soon began building their campus in Phoenix, Arizona. They paid for preview trips there for engineers and managers to consider moving from California. We learned that Phoenix was hot. Very hot. A fellow engineer, who became, in fact, a roommate for a time, concluded, “It makes as much sense to live on the bottom of the ocean. It is equally inhospitable to life.” He also taught me that when you have a rental car on the company ticket, you can wreck it with no consequences. So drive fast, crazy, and drunk. (Bad advice. But he was just kidding. Mostly.)
We asked our much-loved manager, Gene Hill, whether he would be moving to Phoenix. He said, “You shouldn’t let that influence whether the move is right for you.” He lied. Your immediate manager is the greatest single factor in whether you will enjoy your job or be miserable in it.
Gene did not move to Phoenix (I didn’t either) and soon went on to head the first Pentium design team. It was Intel’s first large design team, I believe 100 engineers, as compared to a half dozen on our microcontroller projects. Gene is the nicest guy in the world, but they say the pressures of the project destroyed his marriage.
After four years, I felt I was designing things that I didn’t understand at all, like computers. So I changed jobs and started at Tandem Computers, an HP spinout famous for Friday afternoon “beer busts.”
Tandem wanted me because they were just starting to design custom chips for their machines so my Intel experience was valuable. As a pilot project, they would do a disk drive controller using gate array technology. (Gate arrays are semi-custom chips where only the metal layers are customized.)
By the time I arrived on site, they had decided it was too risky to mess with the disk subsystem so we would do a tape drive controller instead. I was condemned to spend my next six years in disreputable field of magnetic tape technology rather than highly respected field of magnetic disk technology. And, hence, I was unable to ever land a job at one of the big disk manufacturers like Seagate or Western Digital.
I don’t know if it is still true, but they used to say that the highest bandwidth data transfer technology was a 747 aircraft stuffed with reels of magnetic tape.
To place this in time, Al (“Never name a company after yourself”) Shugart had just left Shugart Technologies to found Seagate in Scott’s Valley, which is squeezed between Silicon Valley and the Santa Cruz beach community. Seagate had astonished the world by announcing the first 5–1/4" disk drive with an unbelievable 5 megabytes capacity. That’s right, 5 megabytes! (I wanted to write 1GB, but no, it was just 5 MB.) Floppy disks, which Shugart Technolgies made, were just transitioning from 5–1/4" to 3–1/2".
As I mentioned before, EDA design tools weren’t yet available to do the heavy lifting when I started at Intel, but Intel began to design its own EDA systems. Tandem Computers also had assembled a team in house to design their EDA systems. In 1981, engineers Dave Stamm and Areyeh Finegold, who I knew slightly at Intel, took a bunch of Intel engineers and founded Daisy Systems, the first independent EDA company. It was a crazy rocket of a start-up, monumentally successful. It was eventually acquired by a more successful competitor, Mentor Graphics, still a leader in the field today.
I was a little bit miffed that they didn’t take me, but Dave laughed when I told him and said, “You should have spoken up! We would gladly have had you!” But I was shy little thing.
One of that initial group of Daisy engineers was Steve Sample, who I had worked with directly at Intel. We nicknamed him “Mr. Excitement,” teasingly, because he was so soft-spoken and even tempered. (Sorry to mention it, Steve, if you are reading this.) In 1987, he spun out of Daisy to found Quickturn Design Systems, for which he recruited me out of Tandem Computers.
* long for Medium
Coming up next, Part 2: Start-up Mania.
— Lannie Rose (legally, Elaine Rhodes, formerly, Edward Rhodes)
June 2023
preferred pronouns: she/her/hers
GPT-4 (bing.com/new) used heavily for research, but not at all for writing.
