LONDON – Analysts are trying to think through the implications for global manufacturing of the lost and damaged production of materials, equipment and ICs in Japan as a result of the devastating March 11 earthquake. Despite differences of opinion, it is becoming increasingly clear that the world will now be asked to pay a price for the past globalization of the electronics industry.
Of course, the industry will not be paying so high a price as the thousands of people who lost their lives one week ago or the ongoing and mounting price that hundreds of thousands of displaced Japanese are paying. Our thoughts are with them.
EE Times will publish a special digital edition on March 28 examining the impact of the Japanese earthquake on the island nation and the global electronics industry. All profits will be donated to earthquake relief through the American Red Cross.
But the issues identified around the supply of raw silicon wafers, of chemically engineered resin for packaging, of microcontroller ICs and other issues, highlight the fact that the single-minded pursuit of the leanest, meanest, just-in-time supply chain over the last decade has now put the global economy at some risk.
It is still not possible to know exactly how all this will play out. Some have said that the loss of Japanese production in 2011 will be made up for by the additional stimulus to the economy of rebuilding efforts after the quake. Market researcher IHS iSuppli had warned about an inventory build-up in the supply chain that was happening prior to the earthquake. It now asserts that the inventory buffer could mitigate the worst effects of lost Japanese production. But it is still likely that on one account or another, a lack of adequate second sourcing and disposable manufacturing capacity is going to hurt global GDP in 2011.
There was a time when electronics was seen as strategic, as in a military sense, and domestic suppliers and second sources were mandatory. That has now been largely forgotten. The pursuit of the highest possible volume at the lowest price created vertical disintegration, caused companies to leave markets unless they could be a market leader while shedding operations where they did not have economies of scale.
The result is that the semiconductor industry has become a constellation of specialists that are all highly dependent upon each other. But when one is stricken, there is no one able to take up the strain.
The adage, "For want of a nail," comes to mind. We must now anticipate that an automobile may not be made in Germany for lack of a chemical, a gas, a film or an IC that was still stuck in the Japanese manufacturing supply chain last Friday.
According to Nomura between 90 and 100 percent of the world's supply of bismaleimide triazine is made in northeast Japan. Two Mitsubishi Gas & Chemical plants are responsible for most of the production and they are both closed.
As you say, the question is how quickly other vendors can produce suitable resin and get it tested and approved.
What the industry does NOT want is "nearly identical" materials pressed into service which then lead to chip failures in the field.
@PeterClarke: timely article. There have been mentions made about looming shortage of BT substrates in about 45 days.
I have used BT substrates (common for many multichip packages using BGA & LGA) in many projects that were sourced from China & Taiwan. I do not know the exact % of Japan's share in this market but there are many vendors in the world that can fill the gap. Japan's include Mitsubishi Gas and Chemical, Hitachi Chemical, Sumitomo Bakelite, Nippon CMK, etc.
More than the loss of one country's (Japan's) source of BT, I suspect the supply chain logistics & disruption as the main reason for the predicted shortage of BT. I believe it will take some time for other countries to step up the production and also to get setup in the supply chain. The QA departments will need to speed up vendor audits and placing qualified ones on the approved vendor lists (AVL). Japanese companies have a long reputation of good quality component sources, so the alternates will have to match that.
Dr. MP Divakar
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.