Will the semiconductor “boom and bust” depend on talent instead of sales?

The semiconductor industry is on its biggest roller coaster ride. And in the worst chip crisis, we have also seen some of the greatest innovations of recent years. But while we focus on output and sales, the big risk for the industry is the availability of talent to keep the innovations going.

Will the semiconductor “boom and bust” depend on talent instead of sales?
Talent shortage will soon by a bigger problem than cyclical markets

Everyone in the semiconductor industry can agree that the past two years have been nothing like what we have ever seen before. It has always been one of the most volatile industries, with periods of soaring demand followed by periods of drought. But the laws to which we have become accustomed for so long seem shattered. In the worst supply chain crisis, we have seen some of the greatest innovations of recent years - and the advancements and enablement that memory technologies bring are leading the whole semiconductor industry.

This year, Samsung has started manufacturing logic chips on a 3nm processing node and is planning to reduce the footprint to a 1.4nm processing node by 2027.

If you cannot go physically smaller, the other option for increasing storage density is to go up. That’s what we saw in NAND flash memory this year. A couple of days after Micron announced the shipping of its first 232-layer 3D NAND, SK Hynix announced the development of a 238-layer, 4D NAND flash memory. Following the example of 3D NAND memory, researchers are exploring options for DRAM to stretch upwards vertically to save valuable planar space.

Although DRAM and NAND flash memory continue to advance, every improvement is a painful move that opens doors for emerging memories like FRAM or MRAM. These non-volatile memory technologies are anticipated to finally have their breakthrough with Tech Insider predicting a market volume of $36 billion by 2030. They have the potential to step in for DRAM and NAND as storage-class memory that is faster than DRAM and non-volatile like NAND.

And while there is ominous news about the next bust cycle, it is also clear that the digitization that we have seen spurring over the past two years, will continue and it is not possible without semiconductors and especially memory components. So the overall trend is and remains positive.

Still we should never forget: Behind all these innovations are people. And the big risk for the industry is the ability to build the talent it needs to keep the innovations going. The average age of an electrical engineer in the U.S. is 44 years, in Europe it might even be slightly higher. Add to this that the number of degrees in electrical engineering awarded in the U.S. in 2020-2021 declined by 6.4% with a little over 28.000 degrees awarded.  At the same time, the projected demand for electronics engineers by 2026 is increasing by 6.4%. Although Taiwan has about 20.000 graduates every year, despite a significantly smaller overall population, the demand in the region will also remain consistently high. So it’s very clear that we are going to face a fierce war for talent.

We are at an interesting threshold in technological advancements. But in the end, it will all come down to whether we can get the talent to keep-up the level of innovation and continue to find creative ways to advance it further.

by Marco Mezger, COO and President of Neumonda