Advances in Technology Are Causing Chaos in the Supply Chain

By: SiliconExpert on February 13th, 2019

Guest Blog by George Karalias, Industry Executive Consultant, Karalias Productions

Back in 1965, Gordon Moore, the co-founder of Fairchild Semiconductor and Intel, wrote a paper in which he introduced his observation about the future of integrated circuits. This observation has come to be known as Moore’s Law. Moore predicted that the number of transistors in a dense integrated circuit would double about every two years. For the most part, this has held true for more than 50 years.

Technology advances in the designs of semiconductors are driven by the demands of manufacturers, whose objectives are, in turn, driven by the never-ending demands of a continually expanding number of technology-crazed consumers around the world. It’s the public appetite that’s fueling Moore’s law.

But there’s a catch. Engineers who have designed in components to systems that have long approval cycles and even longer product use life, like automobiles, aircraft, and medical devices, find themselves in the chaotic situation of not having the parts to build their company’s products. Even less-sophisticated products that are manufactured without change for more than several years become victims to this chaos.

Add to this the semiconductor manufacturer’s decision to end-of-life a component because, in their eyes, it is a less-profitable product compared to the newer, upgraded version.

So, what happens when demand for certain components outpaces the supply available? It usually comes down to scrambling to find a suitable form/fit/function replacement or a need to re-design in order to keep the manufacturing of the end product uninterrupted.

However, re-design is not easy. The cost to qualify a semiconductor in a high-reliability application can run into millions of dollars. When originally designed in, the brains of a system (a CPU) requires a line-item-by-line-item programming verification of code. A qualification is then done, first at a component level, then a system level, and then in the actual deliverable item. If a component is obsoleted every three years, the ability to keep production alive over time is severely diminished.

The time required for a re-design comes into the equation as well. Engineering resources are required, and a new design could take up to two years, or more for more advanced applications. And this is just in time for the latest end-of-life notice to start the procedure all over again.

Understanding the potential future availability of a component helps. When designing a product or system, engineers must be highly aware of which components are going to disappear from the supply chain. Yes, there is a need for included technology to be the latest and greatest with all the capability needed for the system. However, having advance knowledge of the lineage of all components and an estimation of what will happen in the near and long term of the lives of those components could save money and angst in the long run.