@ Supply Chain Management

In this part of The New Economics of Semiconductor Manufacturing, an article that you can find at the IEEE Spectrum site, I mean to look at the specific claims and experiences that the authors of the article had in their TPS consulting journey at the unnamed IDM. The following are the four specific "distilled" rules (that earlier Harvard researchers have conducted research on – Decoding the DNA of the Toyota Production System. My own foray into this matter is in this review – Toyota, Toyota

The New Economics of Semiconductor Manufacturing is a new-ish article that you can find at the IEEE Spectrum site which delves into the application of Lean methodology into the wafer fab (fabrication facility) of as yet unnamed integrated device manufacturer. The authors of this article are Clayton M. Christensen, Steven King, Matt Verlinden, and Woodward Yang. Since, I work in semiconductor scheduling and optimization, the experience recounted in this article is of direct interest to me simply because most of these problems have been solved using the "system" I’ve been a part of deploying. Nevertheless for some/many of you who have never walked into a wafer fab, the following is a good snapshot of what you’d normally see:

Walk into a multibillion-dollar chip-fabrication plant-a fab-and you may very well get the impression that the industry is headed for a spectacular meltdown. One of the first things you’ll see is a bay the size of two basketball courts packed with equipment for projecting a lithographic design onto wafers. Nearby, you’ll find a towering bin, called a stocker, filled with wafers waiting to be processed by this equipment. The wafers are worth from US $10 million to $100 million-all of it idle inventory.

Why? To amortize the $5 billion investment in a fab over a five-year schedule costs more than $3 million a day. Conventional wisdom holds that to generate that much money you must keep all the equipment running all the time, even if that means creating large unused queues of wafers. What’s more, to justify that scale, you have to produce a semiconductor product in volumes of at least 5000 to 10 000 wafers per month.

So you can well imagine the wheels spinning in the heads of accountants and finance folk that are dead set on amortizing (all pun intended, "mort" is from the french verb to kill or die) the capital spent in purchasing these very expensive machinery. But that is very much the truth of the matter in my experience where even operations folk blindly follow the utilization mantra in some form or the other. But you should also know that there’s something that is called Moore’s Law

Moore’s law describes an important trend in the history of computer hardware: that the number of transistors that can be inexpensively placed on an integrated circuit is increasing exponentially, doubling approximately every two years. The observation was first made by Intel co-founder Gordon E. Moore in a 1965 paper. The trend has continued for more than half a century and is not expected to stop for another decade at least and perhaps much longer.

While Moore’s law makes this important observation about the rate at which transistors can be "inexpensively" placed on an integrated circuit, it has become some sort of mantra in itself,

Although Moore’s law was initially made in the form of an observation and forecast, the more widely it became accepted, the more it served as a goal for an entire industry. This drove both marketing and engineering departments of semiconductor manufacturers to focus enormous energy aiming for the specified increase in processing power that it was presumed one or more of their competitors would soon actually attain. In this regard, it can be viewed as a self-fulfilling prophecy.

The direct implication of this widely adopted mantra is that the inventory referred to above which is sitting in stockers and in WIP (work in process inventory) any given day is at high risk of obsolescence simply because of development and innovation going on both inside and outside the firm. But also,

More than anything else, Moore’s Law has been responsible for the gigantic costs. It takes huge amounts of capital to support the incessant cycles of investment and obsolescence that keep Moore’s Law on the march. That rapid cycling explains why a company’s shining jewels can turn into white elephants in just five years.

A contrary view of Moore’s Law by Ilkka Tuomi is available here : The Lives and Death of Moore’s Law (and verily to my delight) that follows a strictly empirical approach of analyzing and validating Moore’s law in its multiple variants. I strongly encourage reading this article as it will shed a clarifying eye on the nature of abstractions that we make in this technological march forward.

So regardless of whether you buy the argument that Moore’s law is a self fulfilling prophecy which creates several layers of obsolescence within short periods of time or that development and innovation within the semiconductor industry has largely been the result of interactions between other industries and itself, the facts of the industry do not change in the sense that equipment is costly to acquire, requires significant investments of time, effort and money to use effectively and is a victim/creator of boom-bust cycles. The central claim of the authors is as follows:

In early 2007, we had the opportunity not merely to emulate Toyota’s system but to apply its principles to a logic fab belonging to an integrated device manufacturer (IDM). As consultants, we are not at liberty to divulge the company’s name; however, it’s safe to say that the company is highly competitive-that is, it has survived and prospered by pursuing Moore’s Law, always remaining at the forefront in technology and operational excellence. But Moore’s Law was turning this jewel of a fab into a white elephant while the equipment was still relatively new.

In just seven months, the organization was able to reduce the manufacturing cost per wafer by 12 percent and the cycle time-the time it takes to turn a blank silicon wafer into a finished wafer, full of logic chips-by 67 percent. It did all this without investing in new equipment or changing the product design or technical specifications. And this short experiment has exposed only the tip of the iceberg. We believe that these early results point to what we call the new economics of semiconductor manufacturing and that this will have a profound and lasting effect on the industry and create new opportunities for growth.

In the next part of the series, I will go into the claims made here, essential principles of TPS/Lean that have been used and compare and contrast with my own experiences.

Tags: Semiconductor Manufacturing, Moore’s Law, Lives and Deaths of Moore’s Law, Ilkka Tuomi , Lean within Semiconductor Manufacturing, WIP in a wafer fab

I spent ten days in Istanbul from Jun 18th to Jun 27th and this was a vacation that I’m not likely to forget for a long time. And with the recent attack outside the US Consulate in Istanbul(actually Istiniye), I have a couple of lumps in my throat. That’s because I was at the consulate for three days and I saw, met and talked to some of those guards outside the consulate – I was there to get my visa stamped in order to get back into the US. And a few days later, a gun battle right there where people line up outside the consulate.

Istanbul (earlier Constantinople) straddles the Bosphorous which divides Europe from Asia and for ages been a thoroughfare for trade and culture. Thus, it is no accident that it had been coveted by several empires through the ages. What we call today – globalization et al, is another chapter of the same ol’ story in which Istanbul loomed as one of the players of a bygone era. So visiting Istanbul is like taking a pilgrimage into one of the grand narratives that we traverse – that of trade, of cultures, of peoples colliding, of empires formed and lost etc. As Istanbul is today, so will your favorite city be henceforth – what is to be, already was but we add new pages in our efforts – that is all. All I want to say is that it was beautiful.

I used to love flying – ever since I can remember, I wanted to fly, be a pilot of something or the other. I used to check the itinerary to see how many layovers we’d have traveling back and forth from Africa to India – so that I could "experience" the landings and take offs (As an aside, do you know that English is one of the impossible languages – when I was a kid, I always use to say take-offing instead of taking off because ’to land’ expands into ’landing’, ergo, ’take- off’ expands into ’take-offing’). Traveling to and from Istanbul, we flew Iberia which transits through Madrid and as we were sitting at the gate, I was just watching the jets line up one after the other. I calculated that at the Madrid airport, each jet waited on average 17 minutes before being cleared for take-off. That’s 17 minutes of fuel being burnt for nothing – in other words, WASTE. Now, as you can imagine, the Madrid airport was not really that busy. We flew back into JFK, landed pretty much without circling round and round, all around and round… but after touching down, waited over an hour before getting to the gate at JFK. The flight itself was a little less than 8 hours but the time on the ground was a full 1.5 hours. If you were/are at the short end of the stick of these multiple effects rippling around a busy airport such as JFK – it’s no wonder that flying is a royal pain these days.

Since most of my current work is in semiconductor scheduling these days, I was piqued by these delays. I can almost readily see that transferring/translating my work into airport scheduling and taxiing – it’s pretty much some of the same stuff. So, I looked up similar researchers working on this problem:

An optimisation model for Airport Taxi Scheduling

Airport Surface Movement Optimization (MIT)

And some work that is in progress in the real world:

Surface Movement Advisor

As people around the world fly more, the numbers of flights multiply and airplanes become larger, these problems are going to increase in their magnitude and effects. So here’s hoping for better flying days ahead.

Tags: An Istanbul trip, Constantinople, Flying these days, Airport Surface Movement Optimization, Airport Scheduling

Or rather on my plate. I want to draw your attention to these two challenges that are floating out there in cyberspace – prize money being offered for your (my) troubles as well if one contributes something of value:

1. The first one is the older of the two – The Netflix Prize, which is a problem in improving the accuracy of the movie recommendation system that is used by Netflix (in addition to several other ecommerce sites). This is something that I started working on before I got deluged at work but I hope to pick up the pieces again and make a try.

2. The second challenge is being offered by ROADEF, the French Society of Operations Research – which is to use OR techniques in Disruption Management for Commercial Aviation. I’m sure that if you fly these days, you’d appreciate a little succor from such sources and I’m glad that people are being invited to contribute towards solving this problem. Since I’m working in applying optimization to such problems, this is something I definitely will consider working on.

So, any takers?

Tags: Optimization Challenge, Netflix Prize, ROADEF, Disruption Management in Commercial Aviation

SC Digest has an article by Dan Gilmore, editor, on Creating a “Living Supply Chain”, a concept put out by Dr. John Gattorna of the Sydney Business School (Australia) and Cranfield School of Management (UK).
So what is a living supply chain?

In part, because in the end, it isn’t driven by networks and assets and technology, but by people. Somehow, too many of us tend to lose sight of that. “The reality is that it is people who drive the supply chain, both inside and outside your business, not hard assets or technology,” Gattorna writes. “They are in fact living systems, propelled by humans and human behavior.”

Ok. This is also why management is an art and not a science, its why management consultants are paid mega bucks whether or not the strategies that they recommend are finally put into action or fall by the wayside. So what’s new about that?
Most of my recent posts about supply chain collaboration namely, REA, a semantic model for Internet supply chain collaboration and Using Prediction Markets for Collaboration deal with essentially how to structure one’s supply chain around people and ideas around this particular subject.
One of John’s key observations about a dimension of competitive advantage that is begging for exploitation is recounted below,

It is the failure to understand that human element, or indeed the lack any real body of knowledge in this area of supply chain, that is a major force in why some many strategies go unrealized, and many efforts at collaboration produce little value. Harnessing that dimension is the next and only real source of competitive advantage. “If you can understand and correctly apply a more enlightened approach to managing this ‘human factor’ in the supply chain, you’ll discover a primary source of performance improvement. It’s all there for the taking.”

While harnessing the power of people within the supply chain is not the only real source of competitive advantage but it is one of the important sources and should be tapped. Well, then the automatic question is – HOW?

Read the rest of this entry »

In an article titled Learning to manage complexity, Jonathan Byrnes writes about how growing companies succumb to complexity – their own complexity brought about by growth.

“Success often hurts, and even mortally wounds, well-run small businesses.”

In the article, Jonathan recounts a particularly compelling observation by Daryl Wyckoff,

Wyckoff described a rather strange profitability pattern of trucking companies. Both small and large companies were very profitable, but the medium-sized companies were quite unprofitable.

That problem,

Wyckoff found that small trucking companies that were run by strong entrepreneurs, often hands-on managers, did well and grew. They continued to grow and prosper as long as the entrepreneur could see what was happening in the whole company and directly control all the activities. The problem was that as these small successful trucking companies grew, they typically established a network of terminal facilities. At some point, the complexity of this network prevented the entrepreneur from being able to know and personally manage the whole system.

And further more,

Some entrepreneurs figured out that they would have to manage their companies in a different way. They hired strong terminal managers, delegated authority, and managed through planning-and-control systems. These entrepreneurs were able to continue growing their companies and wound up having very profitable large companies. Other entrepreneurs, however, couldn’t let go. They tried to continue managing their growing companies as they always had. Costs went out of control, and profitability plunged. They had to retrench, and once again they found themselves managing small trucking companies.
Because they were strong entrepreneurs and the companies were small again, they were able to regain control of their companies. They once again made their companies prosper and grow, up to the point where they lost control, lost profitability, and had to retrench yet again. And so the cycle continued, causing the “Bermuda Triangle” Wyckoff described.
In later work, Wyckoff and others found the same pattern in the restaurant industry, the hotel industry, and other similar businesses.

This is something that I read about frequently happening in Silicon valley startups as well. Except that in this case because of the influence of VCs (Venture Capitalists) who are quite aware of this managerial competence gap that entrepreneurs and founders often have, they exert considerable influence and power in bringing in managers and technicians who are well qualified to take the firm into the second stage of growth.

Read the rest of this entry »