Semiconductors have made national headlines in the past two years. The COVID-19 pandemic and its domino effect on global supply chains have caused shortages that impacted a multitude of industries. At the same time, frequent mergers and acquisitions of semiconductor manufacturers have worsened one of the industry’s biggest problems—obsolescence.
Semiconductor shortages are impacting consumers’ ability to purchase not only the latest mobile and computer tech, but also automobiles, white goods, and even LED lights. However, despite their importance in the production of everyday items, few consumers can actually explain how semiconductors work.
Semiconductors are devices designed to control the flow of current in electronic equipment. They are made of materials that are neither entirely conducive, nor completely insulating, and are normally created by adding impurities to elements such as silicon and germanium to alter their conductivity. These are then printed and “carved” into complex designs that allow the current to flow in the desired way.
But why are semiconductors so difficult to produce and commercialize?
The answer lies in the complexity of their design and verification process, and the fragility of the end product. The silicon-crystal wafers that form the base of most semiconductors can contain thousands of circuits, which are created through a complex printing process known as photolithography.
The final product is very delicate and its functionality can be impacted by variables such as temperature, vibrations, and static electricity. This spurred the emergence of dedicated packaging and logistics companies to securely pack and ship semiconductors.
The result is that semiconductors rely on a highly specialized supply chain where each node is interdependent. To complicate things, different countries specialize in different stages of production and logistics, meaning the semiconductor supply chains span the entire globe.
For example, China is the leader in material sourcing, the U.S. is unrivaled in the design process, and Taiwan hosts most of the world’s chip foundries, while most of the lithographic printers are produced in the Netherlands.
This means that at the moment, full self-sufficiency is impossible for any country. As a result, disruptions that impact one node of this complex supply chain may easily affect all other nodes globally.
The Increasing Pace of Obsolescence
In 1970, the typical lifecycle of a semiconductor was expected to be about 30 years. By 2014, this was reduced to ten years—a 60% reduction in less than 50 years. This isn’t a huge problem for consumer electronics, with new smartphones and laptops being released every six months or so.
However, the shorter lifecycle of components can pose a significant challenge for industrial machinery that is expected to last for decades. In particular, highly-regulated fields such as automotive, aerospace, defense, and medical device manufacturing are strongly impacted by fast semiconductor obsolescence. In these sectors, using a different component than the one specified in the original design can lead to extensive and time-consuming testing and verification procedures.
But why are semiconductors becoming obsolete faster than in the past?
Partly, this is due to the natural pace of technological progress, which leads semiconductor manufacturers to phase out older models to make space for more efficient and cost-effective alternatives. Also, older models are normally embedded in legacy devices, which can depreciate so much that continuous support by the original equipment manufacturer (OEM) is not convenient.
However, since the start of the COVID-19 pandemic, another phenomenon has contributed to the fast pace of semiconductor obsolescence—a big round of suppliers’ mergers and acquisitions.
For example, in 2021, Chinese-owned Nexperia acquired NWF, the U.K.’s largest chip plant. In the same year, Renesas acquired Dialog Semiconductor, almost simultaneously with ADI acquiring Maxim Integrated.
When this happens, the purchasing company may decide to streamline the product portfolio of the company it acquired, phasing out semiconductors that are less in demand to prioritize the manufacturing of newer models. This trend further increases the volatility of the already fragile semiconductor supply chain.
How to Cope?
When a product reaches its end of life (EOL), OEMs issue a last-time buy (LTB) notice. This usually gives manufacturers between six and 12 months to buy and stockpile components before they are discontinued. However, LTB typically does not guarantee that all manufacturers using that device will be able to place and receive an order in time.
So, what can be done?
First of all, keeping up-to-date with industry changes, mergers, and acquisitions can help manufacturers predict which components are more likely to be discontinued. To help with that, EU Automation regularly announces when popular components are becoming EOL, both in a newsletter and on social media.
It’s also imperative that manufacturers are aware of the obsolescence risk of semiconductors that are embedded in mission-critical equipment, and that they track the life stage and health condition of such machinery with an adequate predictive maintenance program. This will allow them to spot potential failures before they cause irreparable damage, and to order hard-to-find spare parts before anyone else.
A trustworthy supplier that specializes in obsolete spare parts can also be a great resource and can help manufacturers source quality spare parts from a global network of qualified partners.
These strategies will not prevent semiconductors from becoming obsolete, but they might help offset the negative repercussions of obsolescence on business’ bottom lines.