Japan's Outlook: Device and Equipment Market Trends
Research Department, Press Journal, Tokyo, Japan; Translated by InterLingua Linguistic Services Inc., Redondo Beach, Calif. -- Semiconductor International, 7/1/1998
| At a Glance | |||
| |||
The weaker yen in 1997 as compared to 1996 is one cause for the decrease in 1997, but the major cause is that the DRAM market fell into considerable negative growth because of the drop in prices. The DRAM market in Japan totaled 420.45 billion yen ($3.475 billion), a 22.2% decrease over the previous year, and is the major factor for the slump in the semiconductor market. Other memory products, including SRAM, mask ROMs and MPUs, also recorded negative growth, further aggravating the slump in the Japanese market. MPUs, in particular, were forced into negative growth by the downturn of the Japanese PC market growth and by the drop in prices of MPUs.
Another background factor for the drop in semiconductor devices into negative growth is the slump in the sales of electronic equipment. This is due to the decrease of information investment by Japanese companies, which was caused by the deteriorating business environment, and also because of the decrease in consumer consumption after an increase in the consumption tax rate. The deteriorating business environment is caused by the slump in sales because of the decrease in consumption, which was caused by the increase in the consumption tax rate and a decrease or slow growth of family income because of the slumping sales of companies. In other words, the Japanese economy is in a vicious cycle, which is exerting an impact on the electronics equipment market and the semiconductor market.
The ratio of the Japanese market to the global market also dropped from 25.9% in 1996 to 22.7% in 1997 (dollar-based). The ratio of the Japanese market has had the tendency to drop, influenced by the rapid growth in other Asian markets. This trend accelerated in 1997.
Fig.
1.
Forecast of Japanese semiconductor market (up to 1994: WSTS results; 1995
and later: tabulation result and forecast by Press
Journal).
| Fig. 1. Forecast of Japanese semiconductor market (up to 1994: WSTS results; 1995 and later: tabulation result and forecast by Press Journal). |
Top manufacturers in Japanese market
Let us take a look at the top 15 companies in terms of sales in the Japanese market. This ranking is based on the sales in Japan for each semiconductor manufacturer in the world. Our estimates are included here.
At the top is NEC, in the same position as in 1996, who reached 693 billion yen ($5.7 billion), a 14.5% increase over the previous year. NEC is recording constant growth, despite the negative and slowed growth of other Japanese manufacturers. At No. 2 is Hitachi, who jumped from third place in 1996, even though its sales at 468 billion yen ($3.87 billion) are lower than the sales figure in 1996 (9% decrease). Fujitsu comes in at No. 3, who was previously ranked 4th in 1996, with a sales figure of 451.4 billion yen ($3.73 billion), a 9.4% increase over the previous year.
Toshiba is ranked fourth, dropping from second in 1996, with sales of 450 billion yen ($3.72 billion), a 16.7% decrease over the previous year. The ratio of Toshiba for sales of semiconductors in Japan dropped from 60% in 1996 to 50% in 1997. At No. 5 is Matsushita Electric, who held the same ranking in 1996, with sales of 288.8 billion yen ($2.39 billion), an 8.9% increase over the previous year.
| In other words, the Japanese economy is in a vicious cycle, which is exerting an impact on the electronics equipment market and the semiconductor market. |
At No. 6 is Intel, who maintained the same ranking in 1996, reaching sales of 247 billion yen ($2 billion), a 6.7% increase over the previous year. Although the overall sales figure of Intel as a company achieved a growth of 20% or higher, the growth in the Japanese market is slowing down because the price of MPUs has dropped. Texas Instruments (TI) is in seventh place, up from eighth in 1996, with sales of 243.8 billion yen ($2 billion), a 29.8% increase over the previous year. Good conditions for DSP, also in the Japanese market, contributed to this rise in the rankings.
The rest include Mitsubishi, Sharp, Sanyo, Rohm, Sony, Samsung, LG Semicon and Oki in order of rank. It is noteworthy that the major DRAM manufacturers are not doing very well, and the same is true on the global market. All manufacturers whose ratio of DRAM production exceeds 30% remained unchanged or experienced negative growth, with the exception of NEC and Fujitsu. On the other hand, manufacturers who have been channeling energy into devices with high growth potential and those who have expanded product lines in the niche fields are experiencing steady growth. Just like in 1996, the product strategy of each device manufacturer determined their outcome in sales.
DSP and standard cells show steady growth
This section presents an overview of the trends by product. Just as in the global market, the negative growth of MOS memories is noteworthy. MOS memories in general registered a 15.8% decrease over the previous year, and DRAM, mask ROM and EPROM in particular showed concerted negative growth of 22.2%, 16.2% and 21.5%, respectively, over the previous year. Because of this, the ratio of DRAM to all semiconductor devices dropped from 14.2% in 1996 to 10.9% in 1997. On the other hand, flash memories, which saw growth stopped in global markets because of the drop in prices, maintained growth in the Japanese market with a 17.5% increase over the previous year.
In 1996, MOS logic in general showed a steady growth rate of 11.0% over the previous year, but in 1997, growth slowed to a 5.2% increase over the previous year. MPUs especially fell into negative growth, a 9.4% decrease over the previous year, a plunge from the high growth rate of 1996 (38.8% increase). This is largely because of the drop of unit price in the Japanese market and the decreasing growth in the PC market. MCUs were also forced into very slow growth (0.7% increase), and the growth rate of DSP and standard cells also slowed down compared to 1996, although a 20% level of growth was maintained. This is obviously related to the demand of electronic equipment, an application of MOS logic, which did not grow much in the Japanese market.
Other devices, however, showed steady growth. Unlike the slow growth in 1996, linear ICs showed a 9.9% increase over the previous year, and discrete devices also side stepped negative growth in 1996, even though the positive growth is slight (0.4% increase). Semiconductor lasers marked a high growth rate of 21.3% increase, supported by the demand of optical pickups for CD-ROMs.
7 trillion yen market in 2002
As a forecast for 1998, the biggest concern is how the Japanese economy will perform. Gloomy prospects for Japanese business conditions are rampant. The Japanese government decided on a 16 trillion yen ($132.2 billion) urgent economic countermeasure package to stimulate business conditions, but this has not yet caused stock prices to rise. The government is ready to carry out a special tax reduction as a stimulus measure. This may contribute to a temporary uplift in conditions, but it is questionable whether this will lead to a true recovery in business conditions in the current situation where the equipment investments of companies and the desire to consume among the general populace completely cooled because of the deteriorating business environment and the drop in income. If this slump of economic conditions continues, the above-mentioned vicious cycle will continue, which will make it difficult for the Japanese semiconductor market to escape from the slump.
Press Journal foresees that a genuine recovery of economic conditions cannot be expected until at least by the end of 1998. So the growth rate in 1998 is estimated to be low, a 2.4% dollar-based increase over the previous year and a 7.5% yen-based increase.
Fig. 2. Equipment investment by major semiconductor manufacturers in Japan.
After 1999, however, we forecast that a genuine recovery will be experienced, because the impact of the stimulus measures will gradually appear, personal consumption will generally increase and the PC-related market will be active again because of curtail equipment investments. market will be active again because of the releases of "Windows 98'' and the $1000 PC.
In the year 2000 or later, we forecast that the gradual beginnings of the household electronics market, such as digital TVs, will largely contribute to the recovery of the semiconductor market. As for growth rates over the previous year, from 1999 and later, we forecast an increase of 10.6% in 1999, 12.6% in 2000, 12.3% in 2001 and 12.1% in 2002, where sales will reach 6956.72 billion yen ($57.49 billion)
However, for the market to recover as we are forecasting, various conditions must be met. One condition is that effective measures must be promulgated by the Japanese government to stimulate the economy. Not only measures to stimulate domestic consumption linked to a tax reduction, but also creation of new business opportunities by deregulation will be necessary. It is also necessary for electronic equipment manufacturers to develop new products that will re-awaken consumers and pour energy into a new media infrastructure. Digital TV is receiving heated attention in Japan as the trump card, but even for this product, it is urgent that the government and individuals cooperate and organize an infrastructure.
In this way, we forecast that the Japanese market will recover to levels of genuine growth in 1999 and beyond, but the ratio of the Japanese market to the global market will continue to drop each year, because the growth rate of the Japanese market will be lower than that of other areas. We forecast that the ratio of the Japanese market to the global market will gradually drop every year from 22.7% in 1997, reaching 18.0% in 2002, which will be the lowest ratio in four areas (North America, Europe, Japan and Asia) worldwide.
Semiconductor fabrication equipment industry
We will now look at the trend in the semiconductor manufacturing equipment market in Japan. The Japanese market in 1997 was $6.9872 billion, which is a 5.1% decrease over the previous year (838.46 billion yen, a 4.7% increase). Influenced by DRAM manufacturers who controlled equipment investment, the Japanese market ended with negative growth compared to the previous year.
As for 1998, at the moment the situation remains unpredictable. This is due to aggravated profits because of the continual drop in DRAM prices and the influence of the currency crisis in Korea, which will more than likely cause both Japanese and Korean semiconductor manufacturers to curtail equipment investments. Some Japanese semiconductor manufacturers are considering reorganizing the production system, such as instigating considerable decreases in the production of 16Mb DRAMs, heading for a drastic decrease in equipment investments. Press Journal forecasts that the total equipment investment by 30 Japanese semiconductor manufacturers in 1998 will be 1018 billion yen ($8.413 billion), a 16.1% decrease over the previous year (Fig. 2).
300 mm leads growth in 2000
As for what will happen in 1999 and beyond, the key will be the 300 mm wafer supported equipment, where preparations for the actual installation of this equipment is becoming more serious. The equipment investment scale for 300 mm wafer supported manufacturing lines will far exceed that of the 200 mm wafers; hence, if investments for installation actually begin, growth will again be seen in the semiconductor manufacturing equipment market.
Press Journal performed simulations on equipment investment to construct 256Mb DRAM mass production lines for 300 mm wafers. If it is assumed that throughput in front-end processes is 25,000 wafer per month, equipment prices are ~1.3 times those of equipment for 200 mm and the number of work processes in the front-end processes is 221, then the total equipment investment in an integrated factory will reach 219.93 billion yen ($1.817 billion), which is ~1.7 times that of the 64Mb DRAM manufacturing lines for the 200 mm wafers. Therefore, if investments for the 300 mm lines become active, it will definitely contribute to revitalizing the semiconductor equipment market.
Progress of equipment development and evaluation
The issue here is when this will occur. When will Japanese device manufacturers begin investments for the 300 mm lines? According to a questionnaire-based survey that Semiconductor Latest Technologies (Selete) conducted for 11 semiconductor manufacturers who are participating in the 300 mm wafer program of Selete, the construction of pilot lines will begin in the third quarter of 1998 and will reach full scale in 1999. For mass production lines, some manufacturers will begin in 1999, but most are scheduling a start in the year 2000 or later.
According to our survey, however, the start of mass production lines differs depending on the manufacturer. Some are scheduling a start in 1999, some in 2001 or later and some, depending on the progress of equipment evaluation, are showing a more cautious attitude.
There are still concerns as to whether each manufacturer will be able to start up mass production lines for 300 mm wafers according to the above schedules.
One concern is the economic conditions in the respective geographic area. This is particularly serious for Japan and Korea. When will manufacturers determine the construction of 300 mm lines that will require an enormous investment, more than 200 billion yen ($1.65 billion) per line? Unless the environment for investment takes a turn for the better, manufacturers in both Japan and Korea cannot help but postpone investment schedules for the 300 mm lines. It seems that the current investment priority is to shrink chips by miniaturization rather than to install 300 mm lines throughout the globe.
Another concern is the progress state of development and evaluation of 300 mm supported equipment.
For the evaluation of 300 mm supported manufacturing equipment, Selete, which was established by equal investment from 10 Japanese semiconductor manufacturers (Fujitsu, Hitachi, Matsushita, Mitsubishi, NEC, Oki, Sanyo, Sharp, Sony and Toshiba), is in charge. In concrete terms, a fabrication equipment manufacturer brings equipment to Selete for evaluation, or a client (semiconductor manufacturer) requests Selete for an evaluation.
The installation of manufacturing equipment to be evaluated began in October 1996, immediately after the completion of a cleanroom, and a total of almost 30 units have been brought in for evaluation by the end of 1997. In 1998, 30 to 40 units are in constant evaluation, so the total will be 67 to 68 units by the end of 1998.
In July 1997, a KrF excimer laser stepper was installed for evaluation, and the first sample of a device isolation module using this KrF stepper was completed for evaluation. In August, a consignment agreement with Samsung Electronics was concluded, and Samsung began participating in the 300 mm program. In October, the cleanroom began 24-hour operation, and in November, a cleanroom for CMP equipment and ion implantation equipment was completed, which Selete is renting on the first floor of the Hitachi Production Technology Research Laboratory.
Some Japanese device manufacturers say their decision on when they will begin investments for the 300 mm lines will depend on the progress status of 300 mm supported equipment evaluations by Selete and the International 300 mm Initiative (I300I). Considering the 300 mm line construction schedules that begin in the latter half of 1999 or after 2000, 300 mm line supported equipment must be completed by 1999. According to Selete, however, some supporting manufacturing equipment for 300 mm lines is behind schedule. If manufacturing equipment cannot catch up with the investment schedules of device manufacturers, the recovery of the equipment market will be uncertain. To see recovery in the equipment market, the development schedules of equipment manufacturers must go forward, and needless to say, quick equipment evaluation operations along with schedules is imperative.
A precondition for the recovery of equipment investment by semiconductor manufacturers is growth in the application market, such as PCs and digital TVs, which lead the semiconductor market. Growth of the PC market depends on whether new products, which arouse consumer interest, will continue to appear. The deciding factor will be to what degree can $1000 PCs, which are gaining attention in the United States, and "Windows 98'' lead the market. Digital TVs, on the other hand, are expected to undergo a tremendous boom, but digital TVs and the progress of organizing an infrastructure are indivisible.
Based on these circumstances, we have forecast the trend of constructing 300 mm supported mass production lines and the transition of the fabrication equipment market. The construction of 300 mm supported mass production lines will gradually become active from the year 2000, reaching a peak at around 2002. The speed of actual construction will differ depending on the device manufacturer, but it is without question that the fever of equipment investment by device manufacturers in general will dramatically increase.
Press Journal forecasts that the equipment investment of Japanese semiconductor manufacturers will return to growth in 1999 or later, and investment for 300 mm lines will be active in the year 2000 to 2001. Along with this recovery, the Japanese manufacturing equipment market will also return to growth.