Why Are Some Companies Still Short Semiconductor Chips When Others Are Swimming In Them?

There have been many reports recently of softness in the semiconductor chip market. Taiwan Semiconductor Manufacturing Company (TSMC) reportedly asked its staff to take vacations, personal computer manufacturers have been lowering their chip demand, and demand from smartphone makers has dropped sharply and is forecasted to get worse going into 2023. While some chip manufacturers appear to be swimming (drowning) in inventories, some chips continue to be in short supply. How do we sort out what is going on? The key is to understand that there are very distinctive market segments, and the supply-demand dynamics are different depending on where you sit.

The high end is seeing softness

When we talk about high-end chips, we usually are referring to the newest and most advanced technologies. Semiconductor manufacturers often talk about “nodes” which have a number associated with them. That number historically mapped to the size of one of the critical features of a chip as measured in nanometers (abbreviated nm); one nanometer is one billionth of a meter. The smaller the number, the more advanced. Thus Apple is using chips manufactured on TSMC’s 5 nm node for its latest iPhones, and DigiTimes reported that they are already working on 2 nm. The big consumers of TSMC’s high end chips are Apple, AMD, IntelINTC
, MediaTek, Nvidia, and QualcommQCOM
, and DigiTimes also reported that all are being “conservative” about their immediate sales prospects and hence orders.

This should not come as a surprise. A significant chunk of the volume in this sector is driven by personal computer sales. Back in April 2021 GartnerIT
reported that worldwide PC ships grew the most on record. Chromebooks grew by triple digits then, and all of the top six PC vendors grew in the double digits. But the replacement cycle on PCs is three years or longer, so it was never realistic to think demand would keep growing up to the sky. Maybe some vendors thought this sales pace would keep up, and that’s probably connected to why I saw Chromebooks on sale for $79 recently. Really? The parts are worth more than that! I imagine that warehouses somewhere stuffed full of Chromebooks and other PCs might be related to a drop in demand for chips from this important sector.

Commodity segments like memory appear to be swimming in inventory

Micron TechnologyMU
reported in their last earnings call that they expected supply growth to be significantly above demand growth in the calendar year 2022, and that the sharp decline in near-term demand meant very high inventories in the hands of suppliers. Matching supply with demand in long cycle businesses is always challenging. That’s because supply shortages and strong pricing bring on the animal spirits and new investments in manufacturing capacity, but it takes years to build and equip new facilities. By the time everyone’s capacity comes on stream, there is too much and pricing collapses. PCs tend to use large quantities of memory chips in a single device, so there’s a multiplier effect. Not a happy picture. We see this problem in many other sectors: the flat panel display business goes through what John Matthews described in 2005 as the “Crystal Cycle,” an up-and-down pattern of investment followed by excess capacity.

But the automotive sector continues to struggle

Most automotive semiconductors are produced on “mature” nodes. A recent McKinsey report stated that most of the demand in this sector was for 90 nm and above. For perspective, the 90 nm node was the bleeding edge of chip technology around 2002, twenty years ago. Partly this was because the types of components used in vehicles didn’t benefit from newer technologies, and there is a long and expensive process to move them to more modern nodes. The factories making these chips use older tools, and since these are not terribly profitable commodity parts, there has been little incentive to invest in growing the capacity. Capacity was already tight before the pandemic, and then during a roughly eighty-week period in 2020 when car factories were mostly shut down, most of the OEMs pulled back their orders. Meanwhile an explosion in demand from other sectors filled all that manufacturing capacity, so when the automakers came back to reorder, the lead times had extended way out. They are still catching up.

Aggregation of segments masks underlying dynamics

One lesson we can learn from this mess is that when you take an aggregated view of the semiconductor sector, combining the high end, commodity, and mature chip sectors, it can mask what is going on in different segments of the market. I find the same problem when combing through government import/export trade data or Bureau of Labor Statistics stuff. There’s always great value in understanding the detail.