WAVE Report

Electronica China 2005
By John Latta, WAVE 0517 4/29/05

Shanghai, China
March 15 - 17, 2005

The Tiger Stirs

Electronica & Productonica is meant to examine trends in laser and electro optics in China. Yet, when we got to the Shanghai New International Expo Center (SNIEC), it was obvious that something else was going on. Running in conjunction with Electronica and Productronica China 2005 is the annual Semicon China 2005 event - they are even sharing the same aisles. There were two other trade events in the SNIEC at the same time. The place was hopping. We have not seen such excitement at a trade show venue in years. China has awakened.

This is the 4th International Trade Fair for Components, Assemblies and Photonic Technologies. It occupies one and a half of the long buildings of the SNIEC. The other three and a half are taken up by the Semicon show. This is the semiconductor equipment trade event. Next week is the FPD event run by Semicon. The Semicon show, which opened this morning, was alive with attendees and exhibitors.

China Builds a Semiconductor Infrastructure

The WAVE, having been to three prior CeBIT Asia events and two at the SNIEC, knows the venue well. Under the CeBIT banner, it tried and it was not clear that this show was well targeted. Yet, since our last visit, two new halls have been added – making a total of seven. They were all full this week.

Walking into the Semicon exhibition the isles were packed both with exhibitors and Chinese. These were not the usual tire-kicking public but what appeared to be serious individuals. Many of the exhibits were large and very well manned. Applied Materials had a large booth, for example. Even Intel was here selling its older fab equipment. We noted many American companies with considerable staffs. Represented here is the semiconductor ecosystem. Another tip off was the large containers housed between the exhibit halls. These obviously were used to transport the large booths by ship. This was an expensive and very well staffed exhibit.

Our first response was – is this real or just a hope by the vendors? With that we had many conversations in the booths. Here is what we uncovered:

Most of attendees said that the Chinese are spending money to build fabs. Some were here for the first time and were cautious. But to those with experience in the market, there was no doubt that a major investment is taking place by the Chinese government to build a semiconductor fab infrastructure. Further, Shanghai will be the place where most of this will happen.

In spite of U.S. export controls on the fabrication geometry, this is not expected to be a major barrier to the construction of the fabs. Such controls do not exist in Europe or Japan. Yet, even with fabs that are one and two generations behind, this is not a major impediment to what the Chinese need and are looking for. They just have no immediate requirement for 65nm fabs.

TSMC had a booth and the comment was made by an American that the Taiwanese will not tell the Chinese their secrets to being efficient in fab operations. But in a number of conversations, this was dismissed. An analogy was drawn with the entry of the Koreans in semiconductors and especially DRAM in the 1980s. There was a high degree of skepticism back then that the Koreans could master this cutting edge fabrication technology. Now they dominate it.

In a pattern similar to the entry of Korea, the Chinese expect that when they buy equipment, it comes with training and support. Put in another way, they want the technology when they buy equipment. As a result, they can eventually own the processes and master the technology. As a further result, companies have a hard time charging for training and support. Given the tough semiconductor market, the Chinese seem to be driving the bargains.

Another point made in booth discussions is that the Chinese are cheap. That is they want the best bargain and will drive to no end to get it. Again, not dissimilar to the Koreans when they entered semiconductor fabrication.

When asked – Is there a need for such a large rise in semiconductor production – we did not get a clear answer. That is, no one knows the depth of the demand in China. Yes, everyone quotes the large population that this is largely a myth in that the rural populations have so little money. Thus, the real potential near term market is about the size of the U.S. – 300m. That is enough to still attract significant investment.

The really important issue is that China is on its way to creating massive infrastructure to support a consumer market which could be like none other in the world. China will not only drive growth in the consumer market but be able to satisfy the demands of the market largely internally. In this context, the balance of payments with the US and Europe shows how addicted these regions have become to the existing manufacturing infrastructure in China. Based on what was seen on the technology side this will only get worse.

The WAVE picked up on the floor that China made an announcement last week that it would make a major investment in the TFT LCD industry. We have been unable to confirm this. As indicated above, there is a Semi FPD conference also at SNIEC next week and this is sure to be discussed. The point of this, if correct, is that such an investment will make the current overinvestment in the Crystal Cycle even worse. Thus, the potential for an even greater glut in LCD output lies in the future and that some of the demand from China will be met by internal production. This is not good news for Samsung, LG and Sharp.

Checking the Consumer Market

Running at the same time was ispo – 1st International Tradeshow for Sports, Fashion and Lifestyle Brands. This is the goods market that North Face and other high-end brands sell to. The booths were first class and the place was busy but not as busy as Semicon. Again we tapped the floor for information.

Yes, China has become a very important market for even high end brands. The target market is for individuals from 28 – 40 – and who own cars. This family or individual is active and spends money.

One of the real challenges is pricing. For example, Shanghai will spend more money for the same product than Shenzhen. Thus, issues such as distribution and pricing remain significant to address as China becomes a high-end fashion and brand conscience market.

China’s Entry in Semiconductor Fabrication

If China is able to execute in creating an ecosystem of semiconductor fabrication it has huge implications in the production of electronics components. We went probing some more.

The focus on semiconductor ecosystem development in China is very high volume, one product, mass production. With this focus, the Chinese fabs can relentlessly drive the prices down to levels not achieved with other fabs. This comes at a disadvantage, however. There is no customer focus and changes in requirements, even if slight, are not acceptable. Notice how this is counter to the foundry fab model in place at TSMC and UMC. There is a close parallel with the automobiles in China, even the Shanghai Buick. One model suits all. Such an approach will suit the early market but we strongly doubt this will be sustainable in the long run, especially if the fabs are to supply outside of China. (Note that this has become the strength of TSMC and UMC.)

The China semiconductor industry has a problem with quality. This is a combination of inexperience and focus. The mentality is that things will work out with time but that today quality is less important than quantity. One of the issues is that profitability forces a relentless attack on yield and this is directly related to quality and its management. At present, these forces are just not present. One supplier of discrete components told us that his Chinese buyers shifted their buying patterns to Chinese suppliers. Then, after poor and inconsistent quality experiences they came back.

Some, in their assessment of the quality issue, went so far as to state this could threaten the Chinese attempts at entry into semiconductor fabrication. That is, if these companies do not get serious about quality, it could place at risk the entry into fabrication. One sound bite was “Will China preserve past its problems?”

The current emphasis in production is for 6” and 8” wafers and for specific circuits. This excludes small volume ASICs, custom designs and discrete items. The latter would exclude most EO devices.

China has an advantage with the pricing of the RMB and the strong push by the government to get into semiconductor fabrication.

In spite of the fact that U.S. and European companies are considering joint projects for fab in China, there is considerable concern of putting too many “eggs in one basket.” That is, China has become the world’s production basket, and to also be dependent on China for device fabrication, they would become too venerable. Many see this as not a good sign, the passage this week of the anti- secession act in China, authorizing force to secure Taiwan back onto China.

Many observed that if China is successful in its semiconductor and possibly a TFT LCD production strategy, this could have a major impact on companies outside of China. Companies in the U.S., Korea and Japan are counting on China to absorb their production. If the Chinese can become the lowest cost producers and supply the local market first, this could deny the market to outsiders. This would create a glut of products and possible exit from the market by major suppliers.

One supplier said that the technology transfer from Taiwan to China has already begun. The reason was rather interesting. Some of the engineering and operations talent in Taiwan find that they have no where to go. That is, middle management positions are occupied with young engineers at TSMC, UMC and others in Taiwan. Thus, upward mobility is at a standstill. This talent is coming to China and readily finds jobs where there talent is needed.

There is a Silicon China map, similar to other such Silicon maps. The difference is that this map is of Shanghai.

One supplier commented that the next 3 – 5 years will be the time when China becomes the center of semiconductor fab activity. It is clear that there are many dynamic forces shaping this market and China has, in this technology, the opportunity not only to supply its own market but to fundamentally change the supply chain.

Early Signs of FPD

Truly Semiconductors of Hong Kong and Guang Dong, China, were showing LCD and OLED panels. The most interesting were the OLED. Currently, they have in production a 1.17” panel, samples on a 1.5” panel and the next is 1.8”. The 1.17” panel is used in phone subdisplays and MP3 players. Pricing is $10 - $13 FOB. They get their materials from Kodak.

Truly did do LCD panels but has discontinued these. The panels were being assembled by them from components obtained from LG.

Jiya Electronics, Shenzhen, is doing flat panel production. They make only STN panels and no TFT LCD.

Pale Opto-electronics Presence

There was only the smallest presence of opto-electronics at these two events. In context it was quite simple: discrete components are not a high priority in China and all efforts in these areas are coming from Taiwan as it migrates production to China. The whole area of opto-electronics was so low at these events that even Agilent did not have any such products in its booth.

Taiwan remains a leader in part due to the technology requirements for CD and DVD players. As we have also seen, Pixart is coming on strong in optical mouse ICs and Unity Opto has secured a production contract from Agilent for mouse sensors. At least for the near term we should look to Taiwan for optoelectronics developments not China.

iSuppli Stiffs WAVE Report

Part of the conference program at electronica and Productionica China was an all day seminar on Flat Information Displays Conference 2005 – China. The WAVE Report sought entry but was denied by iSuppli. Yet, it was clear other media had been accepted and to the best of our ability to determine, these were from the Chinese media. It was not clear what iSuppli had to hide but they did not want the event reported by the WAVE Report.

WAVE Comments

Every trip to China is a learning experience. But this was more than we expected.

We have heard the words – China is a huge market – but we have not seen the clear indications of the impact of this other than the malls and shopping districts in the large cities. One of the first signs is usually the development of infrastructure and there is no more fundamental infrastructure in electronics than production and especially semiconductor production. What we saw at Semicon is that this infrastructure is being built. This could fundamentally change the supply chain in the same way that has happened in Japan, Korea and Singapore. China could well become its own self contained supply and demand market in much the same way that the U.S. was some 30 – 40 years ago.

Building a fab is expensive, complex and experience intensive. Operating a fab is even more experience intensive. The one component missing from China is experience. Quality remains a critical issue as China seeks to rush into fab. Further, the mentality of making semiconductors in China has a cultural component – they are creating only a few products for the mass market and in the process driving costs down to levels others cannot achieve. This is consistent with the production philosophy for other goods including automobiles. As a result, customer support is lacking because this is about responding to requirements. Volume is more important than flexibility to meet buyer needs.

We have no doubt that China will be an important market for fab equipment and that it will develop into a major supplier to the Chinese market. However, joint venture partners are showing increasing concern about the risk of relying exclusively on China for all parts of the food chain from component parts to finished goods.

Yet, cultural and political issues loom large and these, in the end, could do much to determine the role that China plays in the global electronics market. Much was made this week about the anti-secession act just passed which authorizes China to use force to secure Taiwan.

There is this magic glow of reaching an ever growing market of 1.3B. Yet, the realities in China are not so simple. A key issue is that outside suppliers of chips and flat display panels all expect to gain market share in China. Yet, if China is capable of executing its fab strategy, and apparently one also in TFT LCD, this could severely truncate external demand. Many a business plan could be left hurting.