High-tech boom and bust cycles

Posted in Technology trends at 10:00 pm by ducky

Twenty years ago, my then-boss Jim Mikkelson gave me a compelling explanation of why the semiconductor industry has such impressive boom and bust cycles. It has to do with the semiconductor manufacturing process being incredibly sensitive and the design cycles being short and somewhat predictable.Imagine that the semiconductor industry has just started a boom cycle. Orders at the fabs — semiconductor manufacturing plants — has increased dramatically and suddenly (for reasons that I’ll explain later). This means that the fab increases its production. (In fab lingo, the wafer starts — the number of wafers put into the beginning step — goes up.)

Positive feedback loops: quantity vs. yield 

Unfortunately, the fab process is incredibly sensitive to changes in the process. A small change can cause the yield — the percent of chips on a wafer that turn out to function correctly — to go down significantly. (At the fab I worked at, we once lost an entire month’s worth of product because a wafer cleaner was miscalibrated.)

One common response is to start even more wafers through the fab. You might think this is counterproductive, but usually the yield doesn’t go down so much that starting more wafers can’t fix it, it takes a while to diagnose yield problems, and the customers want their chips NOW. Unfortunately, running more material through the fab usually makes yield go down.

Hiring more people and buying more equipment also usually reduces the yield, at least in the short term.

Just as the fab starts to get its yield problems under control and get the new equipment up and running, orders suddenly plummet dramatically (for reasons I’ll explain in a minute). So the fab cuts back on its wafer starts, and that starts a positive feedback loop where the yield keeps going up.

Chip customers

At this point in the cycle, chips are incredibly plentiful and, because the fabs have excess capacity and high yield, ridiculously cheap. At this point in the cycle, Jane Engineer notices that and thinks, “Hmmm, with one of those and four of those, I can make a doohickey that is twice as good as what the competition has and at a quarter of the cost. We could take twenty percent of the market with that!” Jane makes a persuasive argument, and sells either her boss or the venture capitalists on the idea, and her team starts designing.

After about a year and a half, Jane’s team’s doohickey is ready to hit the market, so they ramp up orders for the chips they need. Unfortunately, unbeknownst to Jane, nineteen other companies saw the same opportunities that Jane did, at about the same time Jane did, and their design cycle was also about eighteen months. Thus twenty different teams all start ordering as if they are going to get 20% of the market, so demand skyrockets.

The teams, desperate for chips, start double- and triple-ordering chips in the hopes that that will increase their allotment from the semiconductor companies. (The fabs, however, have no way of knowing that the orders are bogus.) When Jane’s project does manage get chips, they hoard them in case the next shipment is delayed.

At some point, Jane’s management notices that Jane’s product has not, in fact, gotten 20% of the market, and that project is in fact losing money. Jane’s management sends in a hatchetman to clean things up. One of the first things that Ms. Hatchetperson sees is that Jane’s project is double- and triple-ordering product when the chips have become quite available and that there is a growing stockpile of those chips in the company warehouse. So Ms. Hatchetperson cuts all orders dramatically.

Meanwhile, the other nineteen projects come to similar conclusions at a similar time, so all twenty projects stop ordering at the same time, with devastating results for the fab. (This can happen astonishingly quickly: at my fab, we went from mandatory overtime in November 1984 to a two-week shutdown in December 1984.)

To summarize, there are four intertwined reasons why the semiconductor business goes through booms and busts:

  • design cycles for products that use semiconductors are short
  • it takes about the same amount of time for all companies to design a certain type of product
  • semiconductor manufacturing yield is strongly negatively correlated with wafer starts
  • it takes about as long for yield to come under control as it takes for companies to figure out that a product is a bust

Note that these were all true twenty years ago, when I was in the biz. It’s quite possible that yield management has gotten much better as a result of the computer revolution. (I certainly hope so!) But I doubt that the short design time, the similarity of schedules, and the optimism that a pet project will get 20% of the market has changed.


UBC trip report

Posted in Canadian life, University life at 9:41 pm by ducky

My beloved husband and I just got back from Vancouver, British Columbia, where we explored the University of British Columbia as one of the options for a MS in CS. The other option is Stanford, which unquestionably has more prestige, but everyone says that Vancouver is a great place to live, UBC seemed like a reasonable university, and we have family just across the border in Bellingham.

We were very impressed with UBC.

Grad student interaction

One place that I think UBC really shines over Stanford is in interactions with other students. At Stanford, it is quite possible to get a MS degree without ever seeing another student, thanks to their strong distance learning program. Not only that, but apparently a lot of the on-campus students don’t bother actually going to class in person when the lectures are available on the Web.

Furthermore, at Stanford, the MSCS is a short, terminal program. While you can do research, it’s not the normal path. This means that the students are not part of lab groups, and I think therefore much more isolated.

At UBC, by contrast, all first-year masters students, who usually are not yet doing research, are given a desk and a SunRay thin client in a room secured with key-card access. After the first year, MS students start on their research, at which point they get a desk in a lab with other students in that research group.

Graduate students are also woven into departmental governance much more than I could ever imagine at any of the universities that I have known. There is a grad student on every departmental committee, including grad admissions, including faculty recruiting, including even the committee that chose a new department head. Barry Po, a PhD student who was my principal contact, apparently had been and is still very involved in getting the new building constructed — up to and including negotiating prices for computer equipment!

CS Department facilities

The facilities there are outstanding. They are JUST moving into a big new, well-laid out, well-equipped building. UBC is a state school, and the winds of political favor happen to be blowing in the University and Department’s favor right now. (UBC is building lots of residence halls right now, in part because of the upcoming Winter Olympics in Vancouver in 2010.)

I like the usability of the new building much more than I like Gates, Stanford’s CS building. (And it’s not just the name I don’t like.) At Gates, you have quite long sight-lines down the corridors, but the ceilings are at normal heights, so it gives an optical illusion of being cramped. In the UBC building, the ceilings are quite high.

At Gates, the doors don’t have windows in them — not even the ones leading in and out of staircases or classrooms. (I’m always nervous that I will whap someone in the face as I go through or that someone will whap me.) The faculty offices have a thin, tall window next to their doors that IIRC is frosted. That helps a little, but is nothing like at UBC.

In the new building at UBC, the faculty doors are mostly glass. They are frosted for privacy, but you at least get the natural light filtering through. Meanwhile, the labs — where the grad students live — have clear glass doors, so you can see people bustling. It feels alive.

There is a snack-bar-ish area in Gates which you can discover if you walk through the entire building (as I did). However, it isn’t at all obvious. They are also experimenting with a small student lounge — big enough for about three overstuffed chairs and a ping-pong table.

At UBC, there are two different lounges in the new building. There is an undergraduate lounge on the first floor of the new building in a very obvious place. It’s quite large — there are about four sofas and about ten tables, with lots of handsome chairs (with “cs.ubc.ca” carved in the backs), and some vending machines. Then, on the top floor, with an absolutely stunning view of the ocean and mountains, is another lounge that is more for faculty/staff/grad students. (I don’t know if undergrads are restricted, but they probably wouldn’t go up there.) It felt bigger than the undergrad lounge, had a sink, and I bet it’ll have a fridge and a microwave at some point.

At Gates, I have the sense that most people keep their doors closed. (Either that or nobody’s ever there!) In the CS department at UBC, it seemed like people kept their doors open. It was a much friendlier environment. (Note that this is probably in large part because Stanford has such prestige. People probably bother Stanford profs a lot more than UBC profs.)


We stayed at Saint John’s College, a grad student/postdoc dorm that has a few couples’ suites, and that was a good move. We got to meet a number of the students, who gave us good info about the CS department, residences, culture, etc. Saint John’s has a mandatory meal plan, seminars, and various events that happen frequently. For example, next week one of the guys we ate dinner with is going to give a talk on a particular composer.One woman showed us the suite that she and her husband have at Saint John’s; it is very small by the standards we are used to, but livable.

Saint John’s is interesting in that it has a strong international theme to it. Another, similar residence hall called Green College has more interesting architecture and a strong interdisciplinary theme.

There are also two apartment complexes that we could live in on-campus. These are much more stand-alone: no meal service, no seminars, probably minimal socializing.

If we go to UBC, we think our first choice is to live in Green’s or Saint John’s if we can get in.

Trailing spouse

Canada is MUCH more accommodating of trailing spouses than the US is. Not only would there be no problem with Jim living in a student dorm, but there would be zero problem with him working up there. He basically just has to fill out a form.

They want me!

UBC is recruiting me very aggressively. They want me. They say they think I’d be a valuable addition to the department. They are reimbursing us $500 of the cost of this trip, regardless of whether or not we decide to move there. Barry Po has spent a significant amount of time wooing me when he should be preparing for his thesis defense next week. I had a full day of tours, meetings with faculty, etc.

Stanford doesn’t really seem to care. I have the sense that if I went to UBC, Stanford would just offer my position to one of the other highly-qualified students with nary a backwards glance.

They are friendly!

People were genuinely, sincerely, actively nice up there. For example, I was wandering around the halls of the old CS building to get a sense of the place. (I like to look at cartoons and posters on doors, see what flyers are pinned up, etc. to get the feel of a place.) Alan Hu, one of the CS faculty, saw me wandering around and asked if he could help me find something. I explained who I was and what I was doing, and asked him a few questions about the department. (It’s always good to talk to people that weren’t hand-picked to talk to you!) He ended up spending an hour or so in a very engaging conversation with my husband and I comparing Stanford (where he got his PhD) and UBC.


UBC acquitted itself extremely well. Now I need to go see if I can bend the Stanford experience into one that would give me as rich interactions with other students (and, to a lesser extent, faculty) as UBC would.


Why ending marriage discrimination for gays and lesbians doesn't imply recognizing polygamy

Posted in Gay rights, Politics at 10:01 pm by ducky

Some people raise a slippery-slope argument against marriage equalityfor same-sex couples (what the search engines will find under “gay marriage”). “If you argue that same-sex couples have a moral right to equal access to marriage, what moral grounds do you have for denying polygamous couples equal access to marriage?” There are a number of counter-arguments to that, but the one I will focus is that it is just not possible to grant equal rights to polygamous spouses. Current marriage laws codify a symmetric, reciprocal arrangement and it is not administratively possible to make asymmetrical marriage laws that are fair to everyone.


There are a huge number of complications that arise from the asymmetry of a group marriage.

Who has the right to make medical decisions? What if two spouses wanted to disconnect an incapacitated one from life support and two did not? It would make the Terry Schiavo case look straightforward.

  • Suppose that Bob and Carol marry. In community property states, Bob gets half of Carol’s community property and Carol gets half of Bob’s community property. Then Bob marries Alice as well without Carol’s consent. Then what happens to community property? Does Bob still get half of Carol’s and Alice’s, while Carol suddenly, through no choice of her own, now only get a third of Bob’s? That doesn’t seem fair. Does Alice get nothing? That doesn’t seem completely fair either.
  • Bob and Carol have been married for fifty years and retire. After she retires, Carol marries Alice. Who gets what percentage of Carol’s pension when she dies?
  • Ted and Mabel are both veterans. Ted has one wife, while Mabel has twenty-three husbands. Ted’s wife gets various veteran’s benefits, including educational benefits and the right to be buried with Ted in a veteran’s cemetery. What benefits do Mabel’s husbands get? If they each get the same benefits as Ted’s spouse, then Mabel’s household gets more money from taxpayers than Ted does. If Mabel’s husbands each get one-twenty-third of the benefits, then they get less than Ted’s wife does. And how would you bury one-twenty-third of each of Mabel’s husbands with her?
  • Alice marries a man from Ghana and sponsors him for citizenship. Jack marries sixteen women from Venezuela. Can he sponsor them all for citizenship? It isn’t fair if he can; it isn’t fair if he can’t.
  • People are granted immunity from testifying against their spouses and in some cases are not allowed to testify for or against their spouses. Can an entire street gang marry each other to make sure that nobody testifies against anyone else?
  • If Jane, Jack, and Lisa are all married to each other, then issues involving divorce — particularly around children — get exceedingly interesting. Presumably, all three would be legal parents of any children. And if we throw Carol and Bob and Ted and Alice into the mix, the paternity might not be clear. How do you work joint custody among twenty parents? If it’s hard enough to collect child support payments from one parent, how difficult will it be to get child support from nineteen?

Decoupling rights and responsibilities from marriage

One thing people have proposed is to decouple rights from marriage. You would get to designate one person to get your spousal medical insurance, one person to get your pension, one person to get your spousal gym membership, etc.

First, while rights are easy to assign to people, responsibilities are much harder to assign. Far more people will sign up to inherit my assets tax free than will sign up to support me in my old age. I could give Bill Gates the right to make medical decisions on my behalf without his consent; I don’t think I could get away with assigning Bill Gates responsibility for my debts.

Second, I am absolutely certain that being able to assign rights to anybody would lead to abuse of the system. I bet it would take less than a day before people started auctioning their spousal health benefits or citizenship rights. I don’t think that’s what we want either.

You might be able to do something where you have sets of rights and responsibilities that are tied to each other; that you can have my assets when I die only if you also sign up for my debts. Figuring out what to tie to what would be extremely difficult.

First-come rights

One easy way to work around the asymmetry inherent in multiple spouses is to grant all of the rights and responsibilities of civil marriage to the first spouse, with no legal right to subsequent spouses. This would be simple and clean, and ridiculously easy to implement — because it’s the legal system we have right now.


It is not possible to have strict equality for polygamous couples; if you gave each spouse all of the benefits, then that household would get more benefits than a similar two-person household. If you give each spouse a portion of the benefits, then an individual in a polygamous marriage would get less than someone in a two-person marriage.

The best you could hope for would be “sort of equal”, and it would be fiendishly difficult to figure out how to rewrite the legal code to do that. The General Accounting Office found 1049 laws that treated people differently depending on marital status. Each state has several hundred laws. Counties and cities sometimes have laws. Government entities and corporations have regulations. All of those laws assume that a marriage is between two people.

To expand marriage laws to include polygamous relationships would mean going through each and every one of the hundreds of laws. The rights and responsibilities could not be made equal, so every single law would require thought and work and arguments about the most equitable way to handle multiple spouses. It would take years.

By contrast, it only takes replacing the words “husband” and “wife” with “spouse” to end marriage discrimination against gay and lesbian couples. It would be easy.