After a big shortage comes a big consolidation, which is typical for a cyclical market like the electronics components industry, no surprise here. A market that normally grows in the mid to high single-digits, and that goes completely out of synch for two years, must correct, as miraculous extra demand does not grow on any tree I know.
The reasons for the latest shortage were clear for some time, the signs for consolidation, too. It started in the computing and communications industry, which represents ~65% of the global components spend, and the next cycle will start there too, with artificial intelligence’s help. The question is when? The other question is “which and how many road bumps are on the way to regrowth?”
A few numbers: the global semiconductor market (ca. 2/3 of the global electronics components market) will – according to World Semiconductor Trade Statistics (WSTS) contract by 10% to $515 billion this year, only to grow again by 12% to roughly $580 billion in 2024. Regional developments differ in percentages, timing, and growth drivers, but this does not matter for the big picture of a market that is facing unprecedented risks and disruptions. No challenge notwithstanding, by 2030, this market will cross the trillion-dollar line.
The road to that Trillion Dollar Realm will be bumpy, disruptive, chaotic and sometimes horrific beyond the usual expectations. Normally, the industry we are in has its own dynamics and challenges:
- New technologies disrupting the old stuff - AI kicking butt in data centers? It’s happening. According to various market researchers, AI will drive additional chip demand north of $50 billion in less than a few years;
- The ever-growing cost of manufacturing semiconductors at newer, smaller geometries - $30 billion for 1 fab? Not impossible;
- The all-electrified society – renewables energy, e-grid, EV? Will eat up semiconductors like steroids;
- Consolidation of players – it will be surprising if we don’t see more M&A activity in the components space in the coming years, from the aftermath of line-card reshuffling and obsolescence waves;
- The big player problem – as if it weren’t enough that TSMC has a stronghold on sub-10 nm production, TSMC also dominates the chip packaging market. And on the customer side, some players have a bigger semiconductor demand than entire continents and could clean out the supply chain of any commodity, should the new smartphone fly off the shelfs;
- Government funding – smart or not, China, Korea, US, Europe, Japan, Taiwan, and others are pumping gazillions into their purported semiconductor independence. What this will lead to, according to Knometa Research, is over 230 active 300-mm-fabs in 2027, way more than the market can digest.
If you think this is more than enough to make proper supply chain planning hard, please add the bigger societal and political picture:
- We are in the middle of a systemic conflict between a western-dominated world and at best a multipolar world. Even worse, democracy is at widely stake and autocratic elites control everything in their countries, from policy to economy;
- The US and China compete for dominion, so far only economically. However, the impact on globalization as we know it is palpable at every corner of the industry, from the bans of critical technology to the withholding of critical raw materials. A future globalization will look different and eventually must calculate with different economies of scale (disastrous for an industry like semiconductors).
- Raw materials needed for high tech will become much harder to get, either through political action or through a much better self-confidence of countries who own them and will not want to waste them anymore for tips. Scarcity of certain metals will come on top, and even circular economies won’t help here as degradation by recycling is the enemy of performance.
- On climate change, the jury has already gone home, it’s happening right now. The ripple effects will pass through the entire economy, water, food, emissions, and prosperity are all on stage and will dwarf the issues of a tiny little high-tech industry, that potentially adds more to the problem than it admits, and if only through the rebound effect (effectivity gains eaten by more volume growth);
- The lack of a skilled workforce in many advanced economies – either by retirement of elder generations or by lack of interest of the youth in many needed jobs/qualifications could be another tumbling block for the high tech industry – and not only here; people needed for skilled jobs are disappearing left and right. By 2030 electronic engineers will be either 70 years old or off to gardening, new ones cannot be found in the amount needed to keep the ship afloat. China, India and some other emerging countries may have an unfair advantage here in the long-term.
I’ll stop right here. You might say, this guy is pathetically negative and shows no way out or offers no solution how take these things into account for our planning. I just list what is on the table, nothing added (some things may be overlooked). The bigger societal picture is aggravating challenges within the industry for sure and raises the question if you can plan your way out of an increasingly unpredictable tomorrow (whenever tomorrow starts) or if it even makes sense to attempt it.
Here is the silver lining: innovation goes on as we speak, everywhere. New things are invented every day (that’s why I think the Trillion Dollar question for semiconductors is not really a question anymore). Smarter people than me see what I see (and more) and if they are able to channel their egos into the good stuff, we will have a chance to turn the current chaos called human world into something better, where technology is used for the right cause, responsible and sustainable and stays within the boundaries of the “one planet” idea.
Short-term, enjoy the rocky horror show.