Thank you for your consolation

[KopiKat]

When they finally found a buyer for CSM (Chartered Semicon), I read with interest the comments by a Temasek top fella who said CSM was never the govt's idea but rather, that of the "private sector" ie. ST (Singapore Technologies). I wonder who's the major shareholder of ST (not STE but the unlisted one)!

Anyway, I distinctly remembered some 20++ years back, when CSM was created, it seems (I must be getting old and memories becoming fuzzy. Maybe if I'm too free, I shld go chk the Newspaper archives) to have been the govt's initiative as one of their multi prongs to focus on the Semiconductor Industry (one of the areas of the Electronics Industry). In fact, there was a major concerted effort to push into this sector, with many other Foundries being lured in eg. TI, TECH (a JV),... with tax incentives and other grants. The supporting industries were also created, with STATS (by that same ST group, I believe) and UTAC (Inderjit Singh?) being 2 of the bigger Semiconductor Test and Assembly cos.

There was also a major push for IC Design cos. and such courses were launched in our Universities and Polys in great fanfare to provide a ready supply of local engineers. Another ST co., (Tritech Microelectronics), originally part of CSM (if I remembered correctly, before CSM was listed) became quite a major force locally. Many MNCs also had IC Design Centres for a while.

I think they were trying to create the entire food chain.

So, what happened?

CSM was one of the top 3 independent foundries in the world. Sounds impressive? No, there were only 3 big ones and CSM was 3rd, after the Taiwan giants (USMC and TSMC). CSM was often lagging behind by at least one generation in their technologies and I don't think they managed to capture the higher profit margin biz cos' of that. When the China foundries started coming in to compete in their "niche" (older technologies), it's no wonder CSM is often in the red.

As for Test & Assembly, it seems like Singapore cos. are unable to dream big to try to be amongst the World Top cos. and like the CEMs which I posted earlier, both STATS and UTAC were privatised (STATS is back again as STATSCHP) and sold to industry giants.

As for Tritech, they got sued for IP infringement, lost the case and disappeared (can't find any info in wiki, but can find info on their lawsuit from some University/News archives or case studies if you do a search). Many of the MNC ones strangely disappeared one by one (perhaps Tax incentives got used up), but, a few of the stronger MNCs are still around.

So, what's next? I don't know... Even Temasek distanced themselves fm this... I think there's still a small niche here for Electronics Engineers, but you'd need to be amongst the top in your cohorts to get in or have the relevant experience (fm Univ projects). Still, I doubt there're that many job vacancies available here..

[arthur]

I am not from the electronic enginnering side but I felt this conversation really opens up alot of this industry history and recent development to outsiders like myself.

Kudos to you all. Engineers are really pitiful ard this island.
Thats why we are huddling around this investment forum hoping to earn an extra buck or two eh.

[weijian]

Yeah, engineers are such a pitiful bunch, that we need to create a film festival and some short films to romanticise our plight...

http://createtomorrow.com.sg/filmfest

Disclosure: I am an engineer.

[mrEngineer]

After looking at Kopikat contribution, I think I will provide a little technical background knowledge behind the electronic industry so that people here can gel up the different players in the value chain. I may have overlooked some stuffs so please forgive me if I am wrong somewhere.

When people start thinking about electronic products, the first thing that comes to their mind is Apple, Sony, Samsung, Intel etc. Actually each of the products can be further broken down to their components: CPU chips (manufactured by Intel, AMD), RAMs (manufactured by Samsung, Qualcomm), Microcontrollers (manufactured by almost all other players like TSMC, UMC etc). What does Venture or Foxconn does then? They actually assemble various components of the microcontrollers, CPUs and RAMs bought from electronic distributors (like Flextronics or Future electronics) into end product (OEM version) and then sell to the end product players like Apple, Sony, Phillips etc. Sometimes these contract manufacturing would have the plastic injection to form the product and sometimes not because of the technical difficulty (specialized equipments or moulds etc).

So I will not continue to elaborate on CPU chips and RAMs as we already know who are the big players and how it works. Let’s concentrate on microcontrollers where most other players are in. What are your microcontrollers? Think of it this way, the camera component, power management unit, sound card, graphic card etc of your Iphone is a microcontroller. Every microcontroller works quite similar to a PC and has a mainboard and CPU chip but this chip will have smaller but focused functionality than the main chipset like Intel. This chip’s are mainly manufactured by wafer fabs by processing of the silicon disc allowing binary 1010s to work. However, we do not usually see the silicon piece when we open up our CPU to look and that is because it is covered by the packaging. This packaging can be sometimes be done together with the wirebonds (bonding of the chip to the peripheral connections), testing and die cutting. It depends on the service that they offer. Stats Chip PAC should offer only testing services as it is already heavy investment on the testing equipment that will perform 24 hour testing (early life failures and sometimes product reliability) non-stop on hundreds and thousands of chips. Wirebonding itself is also a relatively difficult technology as improper bonding may cause poor reliablity of the microcontroller. In the current technology, we will usually see one chip or tiny wafer die cut connected by wirebonds to the package. Now, there are a lot of hoohas on direct connection from wafer die to the package via the substrate. I forgot what it is called but basically that’s the idea.

Next, further down the chain, we will look at silicon wafers and their manufacturers. The big players are TSMC, UMC and there is where CSM comes in. These wafer fabs are extremely capital intensive both fixed and variable costs. Fixed costs are usually the capital cost of the equipments. One lithography machine can cost up to 10-20 millions depending on the technology. One HTOL machine can cost a few millions bucks and this just used for testing reliability of the wafers. Of course there are many more equipments that I can remember and name but it will be meaningless to name all. One just need to know that there are many processes (usually 200+) and each would require a different equipment (although maybe 70% repeated). Variable costs are like material costs, gases, chemicals etc. There will be large storage tanks for these (sometimes poisonous) gases so one can imagine the fixed cost again. Anyway, most of the wafer fabs buy their wafers substrate (plain smooth mirror-finshed silicon) from the substrate suppliers like Soitec or Siltronic. These wafer substrate manufacturers have their own set of specialized tools to melt silicon and process it into 200mm or 300mm wafer discs. So they incur high costs as well.

With this overview of the value chain, one can apply the similar concept to the CPU chips and RAMs as mentioned earlier. They have similar manufacturing process and steps. The HDD will be somewhat different but also has microcontrollers to tell the needle on how to read or write etc. I am not so familiar with HDD so I cant comment further.

Nonetheless, what is happening to the electronics industry? There are 2 major problems. 1st is monopoly of technology and 2nd is the low cost producer from China. When I say monopoly of technology, I refer to mainly Intel. They have been the forefront of wafer fabrication technology with significant breakthrough on the technology limitations like the recent 32nm gate width. This type of R&D generates great growth for them but stifles the whole world to come up with commercial solutions that may revolutionize the industry. 2nd is low cost producer from China, as Kopikat mentioned, many firms worldwide are unable to compete with China as they have siginificant backing from their government as state owned companies and very low labour cost. If you imagine market leaders like Apple goes to China for their parts because of their components price, how do you think the other players worldwide can compete on? This is why CSM failed.

In my view, the massive consolidation will end one day. The big players like Samsung and Intel will not burn money forever and will one day have significant influence on the market, form a cartel and rise prices like crazy. We have to beg them for the products anyway. By then, whoever that survives will win the battle and let the fat lady start singing.

[violinist]

Complicated industry, but still the basic question, and perhaps the reason why Warren Buffett don't touch technology, is what is the Sustainable Competitive Advantage?

As formidable as a giant like Intel or Microsoft, there is still no guarantee that it won't crumble overnight. And who would have thought that Apple could be making and selling mobile phones, mp3 players, and tablet PCs? And such a wonderful company started by Bill Hewlett and David Packard, who hired their neighbours as production operators, assembling products at their back yard, and proudly build up their company into a multinational corporation, would end up with such a mess today.

I think because electrons move at the speed of light. So electronic companies mimic that nature and have to get ahead of the competition at the speed of light. In doing so, we neglect quality, compatibility, standards, reliability, sustainability, and a whole lot of social costs. To those working in this industry, change is the only constant. You'd find yourself constantly being forced to change, and you have to. It's part of the game. I have long since accepted this mantra.

[yeokiwi]

The predeccessor of the current die manufacturing or what we called LSI, VLSI(Very Large Scale Integration) electronics started much earlier in Singapore history. The key manufacturers in Singapore were TI,HP and National Semiconductor. For those who can remember, the TI plant was at the Bendemeer road and NS should be at Toa Payoh.

In those days, they manufactured simple electronics like NAND gate, NOR gate, 7 segment LED, Op Amps etc.

Due to cost of production, obviously these plants were moved overseas in the 90s.

But, we can see the first mover advantage that was available in Singapore at that time. But, apparently, we did not actually follow up and start an intensive build up in the silicon technology. On hindsight, if we had done that, we probably can build up a more sustainable industry with companies that can rival or bigger than TSMC.
The most probable reason was that we were happily ramping up HDD production since the margin was higher.
When the HDD price dropped, we quickly looked for another industry and started to transit to it.

Singapore is normally moving one step slower in leading technology and although we are always faster than most countries, we can never be no. 1. And so, our people will have to get used to a constantly changing manufacturing landscape in Singapore.(Home Appliances->PC assembly->HDD->Silicon->???)

[Drizzt]

there is so little investment in the engineering front. engineering expertise and innovation is soo important to nation from defence and commercial point of view. look at important companies such as nokia and erikson being a cornerstone for first world nations like finland and sweden. we are taking it the wrong way.

[yeokiwi]

there is so little investment in the engineering front. engineering expertise and innovation is soo important to nation from defence and commercial point of view. look at important companies such as nokia and erikson being a cornerstone for first world nations like finland and sweden. we are taking it the wrong way.

I was talking to a Swedish recently and I was quite amazed by their engineering capabilities.
For a 9 million population, they are able to build submarines(hey, we bought the submarines from Kockums) and Grippen fighter jets. One of the largest engineering companies ABB is a merger of a swedish and swiss companies.

In the commercial front, they have IKEA, Ericsson, volvo, saab, Electrolux.

Lastly, the nobel prizes were established by a swedish, Alfred Nobel.

[arthur]

But we are a nation of property agents and financial advisers where our mainstay of living are flipping properties, asking ppl to buy/sell properties and dishing out fantastic investment advises all for a fee.

That's why recent developments on property stamp duties has so many agents clamouring for hopes that the effect won't be far reaching and hit their underbelts.

Who knows? The only conclusion I draw is the government direction for stupendous property price increase is literally killing the golden goose that lay the golden eggs.

Who is the golden goose? NOT FTs, not foreign investors but us, citizens who have a stake in this piece of land. We, who will raise our arms if this land in danger, is being killed off slowly by this stupid government direction, wait.. its called HDB asset enhancement scheme.

If not for the fall of Alj GRC, we can all say goodbye to having a roof under our heads. Its good the incompetent finally get shaken to wake up. Whether its a too little, too late scenario remains to be seen.

[corydorus]

there is so little investment in the engineering front. engineering expertise and innovation is soo important to nation from defence and commercial point of view. look at important companies such as nokia and erikson being a cornerstone for first world nations like finland and sweden. we are taking it the wrong way.

I was talking to a Swedish recently and I was quite amazed by their engineering capabilities.
For a 9 million population, they are able to build submarines(hey, we bought the submarines from Kockums) and Grippen fighter jets. One of the largest engineering companies ABB is a merger of a swedish and swiss companies.

In the commercial front, they have IKEA, Ericsson, volvo, saab, Electrolux.

Lastly, the nobel prizes were established by a swedish, Alfred Nobel.

There is no short cut to engineering. You cannot have innovation with managers. The Swiss neither need triangle to do this.

The problem is traditionally people valued Manager higher than Technicals. And they cannot stand highly paid people laying around playing with gadgets during working time for a long time with a ECO system from engineers to development to scientists.

Without it. We cannot build good planes. Neither will we have a culture to have a team like Apples.

Since Singapore is led by GLC, and SME growths are pressed down by them locally. The GLC has to be radically change to do them proper.


Cory