TSMC has become “too big to fail”, according to semiconductor analyst house Future Horizons, effectively contributing to around two percent of the total world GDP.
Future Horizons' principal analyst Malcolm Penn said at the IFS2012 event in London that the sheer size of the Taiwanese company lends it some impressive comparisons.
“TSMC is controlling about two percent of the world’s GDP, that makes them bigger than Korea as a country,” he said. “It is a huge concentration of power.”
However, this means that all of the fabless firms which do business with TSMC have firmly put their eggs in one basket, and this is potentially dangerous for the industry as a whole.
“You think that this really is now too big to fail,” Penn said, adding that there is a “huge amount of responsibility” linked to this level of dependence on one firm, even if it is one that he describes as “one of the finest companies in the world”.
With the semiconductor industry the basis for other markets, including in electronics, automotive and industrial applications, not to mention a range of service providers such as telecoms and ISPs, the power that TSMC wields is staggering.
In terms of the semi industry alone, TSMC’s means that it accounts for nearly 60 percent of chip production at the 28nm process.
With the popularity of fabless companies, the idea that TSMC could fail is almost unthinkable, supporting the likes of Qualcomm, Marvell, MediaTek, Nvidia and countless others, along with fablite firms as well. Disruption to its production could cause great upheaval in the supply chain.
According to IHS iSuppli, there are more than 150 fabless semi companies which use foundry services, contributing to $30 billion in annual revenues.
TSMC accounts for the lion’s share of this business, with much of it centred around Taiwan, and along with UMC, the country makes up 67 percent of the world’s foundry production and 24 percent of total semi production.
The area is also susceptible to natural disasters, and while TSMC remained relatively unscathed from the tragic events of the tsunami in Japan, earthquakes have disrupted supply in the past.
Whether the chip industry could weather the storm of any sort of disruption to a firm deemed “too big to fail” would seriously test the strength of the industry and almost all of the supply chain reliant on the fabless approach.
Physical placement of the factories is IMHO more critical than just company structures. If bad management ruins a company, it can relatively easy be torn apart and sold to different other companies, while production goes on. But if a natural disaster (earthquake, flooding, hurricane) hits a wide area, then we are really screwed.
There are interesting scientific studies that research the correlation of efficiency and survivability in various systems. With in creasing efficiency, survivabiliy increases initially, but after a certain point, if you increase efficiency further, survivability starts to decrease again.
Applied to economy, it means that merging small companies yields synergy effects, so increased efficiency means better survivability. But a monster like TSMC might be a good deal beyond that critical point already.
At least not all mainboards are build on the small earthquake-endangered island of Formosa anymore, but the production is shifted to various areas in mainland China, that helps a bit. Or look at the floodings in Thailand last year that halted hard drive production and the resulting price hike impacted overall PC and notebook sales.
Perhaps big companies should be restricted by law in order not to take too much risk that might impact global economy (and national even more). Think a penalty tax for putting too much in one place and vice versa a financial incentive to put it elsewhere.