In the first part of this article, I suggested that the Japan earthquake could be the catalyst the electronics industry needs to overhaul and improve its supply chain. (See: Tech Needs a Healthier Supply Chain, Part 1.) In this second and concluding part, I will look at how the industry arrived at its current situation -- and why major steps must be taken to improve it.
Right now, there is enough work-in-progress inventory and other parts for firms to muddle through. But by the end of April, those parts will have been exhausted. Theoretically, companies have some breathing room to secure alternative suppliers, but this will be easier said than done, given the fact that all sectors of the supply chain were working flat-out, with little spare capacity, before the Japan disaster.
A shortage-driven run on components will deplete the inventory much more quickly. Restarting Japanese production has been made worse by the fact that even factories that weren't directly hit are struggling to maintain their pre-earthquake production levels due to power shortages, rolling blackouts, infrastructure difficulties, and making sure employees have enough food to eat and enough fuel to get to work.
Full-shift operation is impossible right now, especially at Japan's crystal pulling and wafer fab plants. If you cannot pull the crystals to make the wafers, which requires 48 hours of uninterrupted power, then you cannot feed the wafer fabs. This will mean an ongoing period of supply shortage hitting home in two to three months' time. The wafer shortage will affect every fab in the world, not just those in Japan, as will shortages of plastics and other materials used in the back end processing.
Assuming things get back to normal by the June-July time frame, it will then take three months for normal supplies to resume. That means the IC shipment business will return to normal operations in the fourth quarter at the earliest (minus the output from any factories that were damaged beyond repair).
We have already seen OEM plant closures as a result of problems with getting supplies; this situation is likely to worsen as the slow process of recovery begins. Fabs need power -- lots of it, and at a reliable flow -- and no one is going to risk restarting a fab or wafer plant until supplies can be relied upon. With 40 percent and 22 percent of global wafer production and wafer processing, respectively, done in Japan, and with no spare capacity, there is zero chance for this slack to be taken up elsewhere, even once problems with securing raw materials, shipping product, and employee absences are resolved.
So much, then, for the "wafers-are-not-strategic" school of thought punted by the IDMs going fabless (or fab-lite), and the OEMs switching chip suppliers at the drop of a price reduction. So much, too, for a world where supply chain security played second fiddle to balance sheet gerrymandering.
Outsourcing everything -- like cheap debt -- looks great on the balance sheet. That's often the only reason why it's done -- not because it delivers better customer service. Outsourcing is also dangerously superficial. So long as everything works okay, outsourcing seems like the right thing to be doing. But when something goes wrong, the effects are potentially catastrophic, made worse by the fact no one really has any contingency planning any more.
Big companies will fail or be radically restructured as a result of the Japan tragedy. The crisis probably won't bring down whole countries, but it might make governments rethink their strategic dependencies, just as they did with oil in the post-war years. The European Commission is investigating the Key Enabling Technologies (KET), such as micro- and nano-electronics, that will drive the development of goods and services in the future.