Travelling On The Silicon Road

Coming close on the heels of ARM’s CEO, Warren East’s remark that ARM based servers will be out in the market in the next 12 months,  comes news from Marvell – the fabless company aims to supply silicon for ARM-based servers with 40-nm multi core processors that it will ship this year. As per EE Times, Marvell and ARM are working with “multiple Tier 1 companies” to build larger trial deployments to validate ARM as a server platform. Separately, Marvell has opened up its dual-core architecture, drivers and board support package so partners can create server software needed to support the new market initiative”

Opinions about ARM’s foray into server market making a dent in Intel’s server biz vary. ARM’s marketing chief, Ian Drew recently pointed out that “We’re an [intellectual property] company — we learned in last two years that we don’t know everything about everything. On the servers, it will take longer than we imagine because it’s a much more complex problem [than smartbooks or phones]”. ARM is running one of its websites (for the ARM Linux Internet Platform) on a cluster of ARM based chips, part of a handful of experiments to test out the viability of using its chip architecture in severs. They are Marvel based.
 

The server market, long dominated by Intel, has lately seen increasing power concerns. Google’s recent acquisition, Agnilux, was rumored to be making servers with ARM architecture. TI is researching the use of DSPs in servers. Intel’s best hope can lie in working along with companies using its low power Atom chips to build power efficient servers.
 

“East wasn’t willing to name the chip foundries that ARM is working with, however he said that information should tip up in the next 12 months.” Now if one of these foundries happens to be GlobalFoundries, it’d be interesting to see how the server market pans out – Intel’s long term rival, AMD and Global Foundries share the same parent company.

Speaking at the International Electronic s Forum 2010 early this month, TSMC’s CTO, Jack Sun, said that he believed a move to 450mm wafers is important for cost reduction and that is going to happen; he reckons middle of this decade. 

The three biggest capex spenders -TSMC, Samsung and Intel want 450mm. But none have contributed substantial funding – an estimated $25bn - $30bn R&D investment by suppliers. 300mm was funded by the equipment suppliers. The same may not be expected this time as the fact whether the suppliers have recouped their investments is still uncertain. Plus they have their other expensive baggage – transition to new process nodes, TSVs, materials etc. 

Now come to the equipment makers. To improve their competetiveness and therefore increase their chances to be selected by the tier 1 semiconductor companies in their future 450-mm operations, the European semiconductor equipment and materials makers support the move to 450mm. However, large non-European semiconductor equipment manufacturers, including Applied, Novellus, Lam, TEL and others, have publicly slammed the idea of the 450-mm wafer transition  

Prior to 2002 (before 300mm wafers actually took off), semiconductor and equipment revenues were more or less aligned. However, post 2002, semiconductor revenues have continued to diverge from equipment revenues, and the ratio of equipment to semiconductor revenues is at an all time low - a divergence attributed largely to the impact of 300mm wafers. 

So the question remains: Who funds the 450 mm, if at all??    

Q1 has been good for the semiconductor industry. Very good, especially w.r.t the dismal 2009.

With Q1 results ranging from knock out to good and promising, industry analysts (iSuppli, Future Horizons etc.) are also revising up their forecasts. The average forecast points north of 30. iSuppli noted that even though semicon revenues typically declines in the first quarter (compared to forth quarter of previous year), Q1’ 2010 saw chip sales up 1..1% compared with Q4 ’09.  Malcolm Penn of Future Horizons has been quite bullish. While cautioning about the potential for a second dip in the general economy, he maintained that only a monumental disaster of a scale similar to the banking crisis of 2008 could now derail the chip market recovery.

However, companies are still treading warily and hesitant to ramp up capacities – there is not much sign of the supply chain easing up.

Close on the heels of analog chip shortages, comes the LCD driver IC one.
While there is a high demand for LCDs (used in notebooks, desktop PCs and cellular phones), chip makers are reluctant to add capacity – the older, high voltage and older technologies used for making them are considered unprofitable.
Companies are going in for alternate options – Samsung and Toshiba has paired up with Toshiba to start production of a “common-type” driver for 256-color super-twisted nematic LCD drivers and Samsung to produce the complementary “segment-type” driver. NEC, a leading driver supplier, is outsourcing to Sanyo Electric.   

As mentioned by Malcolm Penn of Future Horizons in the IEF 2009, “The ‘A’ word is back on the agenda”,
 

But it is not all smooth sailing for the foundry biz. The semiconductor industry’s capex has hit alarmingly low levels. The normal ratio of capex to sales over the industry’s history is 20%. Last year it was 12% and this year it will be 4%. The industry’s overall capacity is now 14% less than it was in Q3.’08. With the economy and market’s forecasted recovery, foundries will be hard pressed next year to meet the demands.
 

And that is where the “C” word comes in – Collaboration and Consolidation.
 

There has also been a lot of talk on consolidation in the foundry biz – as in other areas of the semiconductor industry. There are some pending mergers - between Hua Hong NEC and Grace Semiconductor; Tower Semiconductor’s 2008 purchase of Jazz Semiconductor, proposed acquisition of HeJian Technologies by UMC and the recent purchase of Chartered by GlobalFoundries.
 

Possible future mergers are: SMIC acquiring Cension Semiconductor Manufacturing International and Wuhan Xinxin Semiconductor Manufacturing - two companies which SMIC is managing. And then there are small foundries like Silterra, Altis and Landshunt which are struggling and open to speculation regarding a merger with another manufacturer.”
In all likelihood, as cited by iSuppli, there may be 3 major pure-play foundries left standing after the consolidation – TSMC, UMC and GlobalFoundries.

.

The solar industry challenged semiconductor equipment manufacturers during the Photon Technology show in Munich in April this year, to better address their needs. While in the past, they had to make do with equipment originally developed for the semicon industry, they are pushing the equipment suppliers to build specifically for them.

Looks like, not only the equipment and yield management solutions/services vendors (like KLA-Tencor, Applied Materials etc.), but also the EDA vendors are jumping on this bandwagon; in the process also attempting to improve their bottom line by leveraging from this rapidly growing industry; especially important in times when the forecasts for the semicon industry do not appear too rosy.

With limited resources and the present state of economy, hope this does not come at the cost of inadequately addressed solutions to UDSM design related challenges.

The story is familiar – Foreign companies come into a country to tap into its low cost labour. Sooner or later the low cost factor erodes and those foreign companies move out to newer pastures.

This is what also happened to the semiconductor industry in China. Foreign electronic equipment manufacturers come to tap into its relatively low cost labour. With the global economy going south, they decelerate the pace of their manufacturing outsourcing to China resulting into negative impact of the country. However, this negative impact has partially been off-set by a transformation in the country wherein there is now an increased focus on designing chips for electronic products that are popular in the nation - and thus tapping its own vast domestic market.

Demand for these locally designed chips is being driven mainly by China and Hong Kong-based electronics OEMs and contract manufacturers with foreign ODMs that develop and manufacture goods for Chinese OEMs also chipping in.

The initial manufacturing outsourcing to China had resulted into a large number of local design houses – they are now tapping the domestic market.

The largest application for Chinese designed semiconductors in 2007 was Mobile handsets followed by notebook PCs. But the fastest growing segment for domestically designed semiconductors over the next few years will be mobile communications infrastructure equipment - not surprising as the country is home to two of teh world’s bigger players i.e. Huawei and ZTE who incidentally are aggresively targetting the international markets for their growth.

Reports that Samsung, the world’s largest buyer of capital equipment, is now putting the brakes on its capital spending, has exacerbated the doom-and-gloom sentiment. Poor memory environment coupled with economic woes have been cited as the main reasons behind semicon equipment vendors’ woes.

A shake-out in the fab tool & electronics materials industries is expected.  While consolidation through acquisitions & mergers loom on the horizon (Applied vying for ASMI, Aquest for Asyst etc.), we also see the semicon fab tool vendors gravitating towards the booming solar industry.  Applied has been diversifying out of chip cap-ex and equipment dominance into a strategy of being a solar player, KLA Tencor also entered the solar market with purchase of ICOS vision systems, a company whose primary business is inspection equipment for semiconductor packaging but which has been successful in the more cost-conscious solar wafer inspection market at key processing steps.

The recent SEMICON West had more than 250 exhibitors that had offerings for both the photovoltaics and the semiconductor markets. But it seemed that nearly every exhibitor had some sort of solar story to tell. From manufacturers of high purity chemicals used for microelectronics selling solar cell surface cleaner to Synopsys providing modeling software for the PV industry, there’s a range of interesting synergies being tapped across the two spaces – recommend you to read Semiconductor International’s write-up on some interesting solar stories from the floor

According to a recent research in Semiconductor DQ report from Gartner, it is noted that the top 100 OEMs consumed semiconductors worth $209 billion in 2007 or a total 76% of semiconductors sold last year. However their semiconductor consumption did not grow as fast as their revenues. The report highlights 2 ongoing issues faced by the semiconductor industry - significant erosion of semiconductor content and average selling prices (ASPs),
 

While ASP’s have long been an issue especially in the computer and cell phone market, the other aspect of decreasing semiconductor content is a worrying trend for the semicon industry. True, we are seeing some signs of new and innovative uses of semiconductors, such as sensors being implemented in products by firms such as Nintendo & interesting new touchscreen-based products, but then that by itself may not be enough.
 

Reverse trends include:

• Talks of Apple selling newer versions of its iPhone s/w through its iTunes stores. Unlike traditional mobile handsets, where users change their handsets quite frequently (hence more semicon content demand), iPhone users may not change their handsets so often and upgrade their phones mostly through relevant s/w upgrades.
  
• Recently there was this interesting news article in Biz section of International Herald Tribune (dt. June 23 ’08). It reports that Nokia wants to transform itself into the next generation entertainment company. It has already created (last August) an Internet service and online music store, Ovi, said to compete directly against Apple.
   Nokia predicts that in the next 5 years, phone users will create 25% of entertainment watched on smartphones. – and that Nokia will share that entertainment.
   

Both these examples indicate a growing dominance of s/w content and potentially diminishing h/w demand in the computing and consumer industry – areas which already show a pricing weakness (Gartner made note that the semiconductor content pricing in these areas decreased almost 9% in 2007 compared with 2006).
 

How does this bode for the semicon industry???   

Not too long back, quite a few companies, mostly ranking in the semicon top 10, counted a diverse portfolio as their strength. With the brutal and dynamic market conditions, this is becoming a luxury that only a few can afford. 

Following the trend of consolidation and focusing on core biz activities, Infineon will sell its hard disk drive (HDD) design and manufacture business to LSI Corp. LSI expects the acquisition will further its goal of becoming the leading worldwide provider of silicon solutions for hard-drive makers, especially at a time when the HDD market is going from mechanical to solid state drives. And they get an inlet to a top-tier customer, Hitachi Global Storage Technologies.

For Infineon, it is shedding a non-core biz, an area where they did not stand much of a chance amongst competitors like Marvell, LSI, TI and STMicroelectronics – and to focus its resources on its core businesses where it can be among the top few.

Just a year back, another acquisition between the same companies took place; albeit in a reverse direction - last August, Infineon agreed to pay $563.6 million to acquire LSI’s mobility products business to strengthen their position in mobile phone market. And the top tier customer base to be tapped there was Samsung.

They sure dig each others’ technology!!