Are semiconductors headed for a meltdown?

Source: This article is from semiwiki, thank you.

In the 2022 Update, Malcolm spent an hour detailing 33 slides, including his previous high-end forecast of 24% semiconductor revenue growth in 2021, which ultimately came in closer to 26%. The most important parts of the presentation were the historical review of the semiconductor industry and his forecast for 2022. Spoiler alert, a crash may be coming.

If you look at his opening slides, you can see ups and downs throughout history, including the 2020 dot-com bubble I mentioned earlier. One of his slides goes into more detail on previous ups and downs since 1961. While the current 2019 bust and 2021 boom doesn’t exactly match up to the dot-com cycle, it’s still significant.

Malcolm also mentioned EDA, but let's look at it in more detail. EDA was also a single-digit growth industry, but recently, as you can read on SemiWiki, EDA growth has been booming with double-digit growth.

There are two reasons for this: First, system companies are scrambling to make their own chips, and this now includes car companies. The chip shortage is a big driver, but the increasing software burden on system companies is a close second. Self-driving cars now contain millions of lines of code that can be developed and optimized in parallel with chip design. Smartphone companies realized this a long time ago when Apple and others started making their own SoC chips.

Secondly, venture capital is pouring into the chip space at a record pace. AI is a big driver for startup chip companies, and e-cars are another bubble waiting to burst. Last told there are 300 companies developing AV/EV related products, I mean WOW. Again, deja vu dot-com bubble.

Malcolm then turns to the key drivers, their impact and role. The key drivers are economics, unit demand, capacity and ASP. Malcolm covered that in detail, but I will make additional comments on capacity.

We have capacity, that has never been a problem. Leveraging that capacity is another matter. For example, TSMC's automotive chips fell by -7% in 2019 and -7% again in 2020. That's why car companies don't have enough chips and cancel orders. In 2021, TSMC's automotive business grew by 51%, and 2022 will probably remain the same because inventories are growing at an unprecedented rate.

But the chip shortage story continues, and so does the capex race between Intel, TSMC, and Samsung. The big difference is that TSMC builds capacity based on customer orders, paying some big bucks up front, while the IDM foundries don’t. TSMC is also building large capacity for Intel, which complicates things a bit. The one saving grace is that it will be hard for equipment companies to equip all of these new factories with the supply chain issues they’re suffering. Especially ASML and EUV. They can’t possibly equip all of the new leading-edge logic and memory factories that are in the press release phase.

Malcolm continues his agenda and takes us close to the top of the roller coaster. He doesn't show how big the drop is, but he's sure it's coming. He also has a nice graphic. His forecast for 2022 is 4% on the low end and 14% on the high end. I'm more optimistic with an industry forecast of 10-15%, with TSMC hitting 20%+ growth rates again.

Malcolm finished up with the key takeaways and Q&A. For me it all relates back to the supply chain that Malcolm detailed. When the dust settles and COVID is under control we will see a much stronger supply chain that will not be re-educated, of course that is just my opinion.