The Beginning of the End of the Automotive Industry Semiconductor Shortage

Our latest quarterly update of the outlook for?automotive?semiconductor?demand has just been released, Automotive Semiconductor Demand Forecast 2020 to 2029 - October 2022. Automotive semiconductor demand will maintain growth with the market forecast to grow by 22.2% to $53.8 billion in 2022, as semiconductor supply continues to play catch-up to vehicle production. This is augmented by higher ASPs and the continued momentum towards electrification underpinning moves towards domain-based and zonal E/E architectures that will drive ADAS/autonomy, advanced infotainment, telematics, and vehicle connectivity.

Over the 5-year period 2021 to 2026 the CAAGR is expected to be +16.3%. Excluding semiconductor-based sensors, the CAAGR for automotive semiconductor dollar demand is expected to be even higher at +16.8%. The global automotive semiconductor market value is forecast to be $93.6 billion by 2026 and reach $111.5 billion by 2029 (including semiconductor-based sensors).

The Beginning of the End of the Automotive Industry Semiconductor Shortage

The ongoing COVID situation and the war in Ukraine, and subsequent fears over the global economy, continue to be the dominant factors in this update, along with ongoing attempts to resolve semiconductor supply issues. Our view that a semiconductor shortage scenario would remain in place for the remainder of 2022 looks like it will be held true, and there is a view amongst some automakers that the situation will prevail through 2023 also.

The confluence of events around COVID-related lockdowns in China, the Ukraine conflict, a slowing global economy and rising inflationary pressures has led to a notable slowing of demand from consumer-focused markets PC, mobile and data centre markets. This in turn is leading to a build-up of inventory for some semiconductor suppliers as end customers adjust their ordering. On the other hand, Strategy Analytics sees no slowing in the push towards electrification, which in turn is underpinning the move towards high-end infotainment systems and broader adoption of ADAS (advanced driver assistance systems) and vehicle connectivity.

However, while the slowing in demand from other sectors will undoubtedly free up capacity to serve the automotive industry, there is no linear correlation between semiconductor demand in the consumer-focused markets and the easing of semiconductor shortages in the automotive industry. A key factor in this respect are the requirements for automotive qualification as well as continued use of older generation process nodes which are not used by the PC, mobile and data centre markets.

To the latter point, the automotive industry has continued to see bottlenecks occurring with components that are based on process nodes older than 28nm, e.g., 40nm etc. In these cases, the fabs have usually not been upgraded so they have filled up very quickly and are running at full capacity presently. There is no overhead and no real incentive to invest to add capacity here since it is expected that these processes will be superseded over the coming years in any case. While supply should have started to catch up to demand, the push towards new platforms coupled with an emphasis on the premium sector significantly raised the overall semiconductor content per vehicle leading to bottlenecks.

These bottlenecks are now starting to ease, with supply of more mature products starting to close the gap with demand. However, requirements from electrification and the associated trends around ADAS/automated driving, infotainment systems, connectivity, domain, and zonal/centralized architectures shifted focus towards higher performance and specialized semiconductor requirements. These requirements were not in place prior to 2021, and this rapid shift in focus continues to represent a supply-demand imbalance that will be worked through in 2023.

As noted previously, the move to domain and zonal architectures is not characterised by a collective standards-based automotive industry effort. So, it will take time for these and other changes to impact the semiconductor supply chain, which is why our forecast model still shows demand for discrete ECUs with their own distinct semiconductor BOMs.

We assume that early iterations of new OEM platforms, while driven by the push towards electrification, will also underpin ICE-based powertrains moving forwards. As such, we will still see continued use of individual discrete ECUs in use for a while. Software enabled capabilities will start to virtualize non-safety-critical ECUs in later iterations of these platforms and this will start to filter through towards the end of the forecast timeframe.

This should also translate to a move away from application specific standard parts to semiconductor suppliers developing parts that can be used across a broad range of applications, subject to the relevant qualification requirements.

While the acceleration away from legacy distributed E/E architecture towards domain- and zonal-based architectures is not in doubt, there will be a long tail of market demand for ICE-based powertrains which will maintain demand for multiple ECUs across the vehicle E/E architecture.

The risk of excess inventory build remains even as most automotive semiconductor suppliers have entered long-term supply arrangements with their customers. In many cases, these are NCNR (non-cancellable, non-returnable) agreements which provide a certain measure of protection. However, the quarterly semiconductor industry financials being released continue to point to robust demand and hence revenues from the automotive sector.

We also expect that the pricing for strategic components that support the move towards electrification as well as associated trends of ADAS/autonomy, connectivity, domain/zonal and centralized architectures will not see a return to the same demands for year-on-year price declines as traditionally observed for mainstream automotive components until at least 2027.?

The supply-demand imbalance will remain through 2023 driven by increasing semiconductor content per vehicle and more advanced semiconductor requirements that have emerged more rapidly than was expected or planned for by most of the legacy automakers. However, Strategy Analytics believes that we are now starting to see the beginning of the end of the semiconductor shortage and a normalization of the supply-demand balance will start to be observed from the second half of 2023, with normal service resumed from 2024.

We’ll be analysing the automotive semiconductor company financials over the coming weeks to see whether the outlook presented in our latest forecast aligns with company revenues and their expected outlook for this year and beyond.??

Feel free to contact me to discuss this post and the underlying questions raised.

Also, check out our latest quarterly view of the global outlook for?automotive?xEV?systems?and associated?semiconductor?and?sensor?demand. Global xEV System, Semiconductor and Sensor Demand Forecast 2020 to 2029 - October 2022 ?shows the continued momentum pushing market demand to $30 billion by 2029.

If you’re?looking for more granularity in the xEV segment, then check out the full suite of?datamodels?presenting?global?and?regional (Brazil, China, East Europe, India, Japan, NAFTA, Russia, South Korea, Thailand, West Europe and ROW) outlooks for xEV systems and associated semiconductor and sensor demand as well as battery technology and charging infrastructure trends from the Electric Vehicles Service (EVS).

Part of TechInsights, Strategy Analytics offers extensive coverage of the automotive industry, through the EVS (Electric Vehicles Service), PBCS (Powertrain, Body, Chassis & Safety), AVS (Autonomous Vehicles Service), AIT (Automotive Infotainment and Telematics) and ACM (Automotive Connected Mobility) services.

@StrategyAnalytics @TechInsights #automotiveindustry #semiconductorshortage #electricvehicles #Ukraine #China #COVID