A Conversation with... Dr. Ping Yang of TSMC, Taiwan
Alexander E. Braun, Associate Editor -- Semiconductor International, 10/1/1998
SI: How is Taiwan responding to the IC industry's general downturn?
Yang: In general, we are doing better than most of the rest of the Asian countries. We have a healthier, stronger foundation. The electronics industry here is not doing too badly, not to say it couldn't be better, though. We're all tightening up our belts a bit. Some companies are asking people to take vacation; their top management has taken cuts in pay, and others are on short four-day workweeks and have gone off three shifts, not too different from what things are like in Silicon Valley, as I understand.
SI: How is morale holding up?
Yang: Morale is pretty good. We Taiwanese tend to be extremely pragmatic. We know that this is a situation we just have to work along with and tough out, and that things will get better. We are very upbeat.
SI: So Taiwan is making lemonade out of today's lemons?
Yang: (Laughing) Indeed! Many of us believe that this is an ideal time to work on the fundamentals, the foundations. When a company finds it has little work to do, its people don't sit around twiddling their thumbs. There are ongoing programs to train personnel and bring everybody up to the levels that will be necessary to meet the challenges of next-generation technologies. Some are not only doing this and planning for the future but are actually carrying out re-engineering, making sure that they will be prepared when the upturn comes.
SI: When do think that will take place?
Yang: We think this will happen during the second or third quarter of 1999, and we're talking about a major upturn! This is pretty much the consensus here. We're seeing things picking up already, during this third and into the fourth quarter. This is not the major upturn we expect; this is only the beginning of the readjustment upward.
SI: Has TSMC or anyone else in Taiwan delayed plans to develop 300 mm wafer processing capabilities?
Yang: Not to my knowledge, but there is a caveat. In TSMC's case, for example, we have not delayed our plans. However, as you know, the equipment vendors have put off the transition to 300 mm by diverting their resources away from 300 mm. Also, SEMATECH is not pushing 300 mm as hard as before. Our position is that we will not delay it, but it may be delayed nevertheless if for no other reason than equipment vendors are hesitant to produce the necessary hardware. One thing we will not be is a guinea pig, meaning we do not want to be the first ones on the block. We will have everything ready, but if the equipment is not, then there will be, of necessity, delays.
SI: So when the time does come to go to 300 mm, would you use your current facilities or build new ones?
Yang: The answer is yes and yes. We are going to build a new facility that will have two fabs; one will be a 200 mm facility, and the other one is slated for 300 mm.
SI: Does this mean then, that TSMC will continue to participate in I300I?
Yang: Definitely!
SI: Do you and other Taiwanese IC manufacturers plan to introduce copper and low-K dielectrics any time soon?
Yang: We already have a lab investigating copper and doing development work. During the third quarter of 1999 we will introduce copper in volume production. Similarly for low-K dielectrics. We have had it in small-sample production already. We are 100% certain that when we go into 0.18 µm production by the first quarter of 1999, everything will be in place. We're already using low-K dielectrics - judiciously - in some limited 0.25 µm device production. Our low-K will be ready for volume production early next year.
Especially for this issue, Semiconductor International's, Alexander Braun interviewed Dr. Ping Yang of TSMC, Taiwan. Yang received his master's and doctorate degrees in electrical engineering from the University of Illinois, Urbana. He has been with Taiwan Semiconductor Manufacturing Co. (TSMC) since August 1997 as vice president of corporate marketing, after working at Texas Instruments, where he joined its Central Research Laboratory, to pursue VLSI research. Yang is an IEEE Fellow and has co-authored two books and numerous articles.
" We're seeing things picking up already, during this third and into the fourth quarter. "
|
Dr. Yang
SI: In the past, Taiwanese companies have entered into a variety of strategic alliances with Japanese, U.S. and European companies. Will this trend continue?
Yang: Yes, fortunately and unfortunately (laughing). I think the reason is very simple; Taiwan is good at manufacturing. At present, this is our greatest strength. In terms of fundamental R&D, however, we have a long way to go as yet, and the same is true in terms of sales and marketing capabilities. Because of this, you will certainly continue seeing the building of these alliances, especially those which complement the partners' various strengths and weaknesses. Also, it is a fact of life that global competition demands these arrangements. Personally, I hope that eventually these will become purely strategic, rather than complementary, alliances.
SI: Taiwan's IC business is largely focused on supplying foundry services and on DRAM manufacture. Are there plans to get into other areas?
Yang: Actually, if you look at companies in Taiwan, foundries and DRAMs are the new phenomena. It is very much like the U.S.A.; IC companies there are mostly fabless. Since they do not have fabs, they have a far superior business model, and they have truly blossomed. The same thing has happened in Taiwan. Most standard-product companies here - logic products and microcontrollers - are blossoming. But you do not see it, because they do not spend large amounts of capital to build fabs. The only exception is DRAM manufacturers, because these cannot really be produced effectively in somebody else's fab. So it's the DRAM companies that must spend all that capital. In a way, Taiwan is a combination of the fabless companies (like in Silicon Valley) and fabs.
SI: What do you see as the role of the Taiwan Center for Semiconductor Research, started by MOEA in June last in your country's R&D efforts?
Yang: TSMC will participate in the center's efforts. We believe that we will benefit somewhat from the effort. The reality however, is that Taiwan has many small companies and a few large ones. These companies are behind the large ones in terms of technology, while the large ones keep pace with the rest of the world. The center's role will be similar to that of SRC and SEMATECH in the early days, and although everybody will benefit, the larger companies will benefit less. It will help Taiwan in terms of R&D; I don't know how much it will help TSMC, however.
SI: What do you see as key areas of leadership required for the future success of the industry in general and TSMC in particular?
Yang: From our viewpoint, the capability to do technology development and design expertise are critical. If you look at the way foundries work, it is similar to the OEM business. DRAM manufacturers license their technology to form strategic alliances, so in a sense they are also somewhat like OEMs. This is not the way to compete in a high-tech environment. We believe that in order to have a solid technology foundation, we must develop an independent design capability.
SI: Any words of advice to your peers and the industry in general?
Yang:
Don't be satisfied with just getting paid to do manufacturing. We must
outgrow that mode. We are very fortunate in Taiwan in that we are very
nimble. The whole industry can switch overnight from something that was
hot before to something that will be hot tomorrow; you must remember
that in our industry technology becomes obsolete every three years. The
DRAM industry has always had that problem mainly due to its heavy
investment in whatever is the current technology and tended to lag
behind. It is to be hoped everyone learns the lesson that you must be
able to adjust quickly and efficiently when technology changes
direction. In Taiwan, we have had the tendency to kill the goose that
lays the golden egg. When people see others making money, they don't
stop to think why it is they are making it; they jump in and use
cutthroat methods to benefit. This is very shortsighted and does not
provide for the future. I just hope the DRAM fiasco has taught everybody
a lesson. 
