Editor’s note: Supplyframe, which was acquired by Siemens in May of 2021, is a market intelligence firm collecting billions of continuous signals of design intent, demand, supply, and risk factors to better inform electronics design and purchasing decisions. Supplyframe is part of Siemens’ Digital Industries and its software-as-a-service (SaaS) offering.
EPSNews contributor Bruce Rayner recently spoke with Richard Barnett, chief marketing officer for Supplyframe, about a range of topics, including automakers’ plans for chip design.
Q: We are two years into the Covid-induced supply chain crisis. What are some of the critical issues OEMs are grappling with now?
RB: “In early 2020, the first wave of Covid was shutting down manufacturing plants in China, then around the world. That initial round of plant closures consumed 60 percent to 70 percent of the total end-to-end inventory buffer. Once it was gone, there was no chance to catch up. At the same time, the demand mix for products changed dramatically. No one was buying cars because everyone was home working or going to school virtually. Instead, audio ICs, assemblies, and submodules destined for shuttered automobile plants were diverted to factories building PlayStations and Xboxes. It wasn’t company versus company. It was industry versus industry competing in a market with growing supply constraints.
“The clock speed of IoT sensors and semiconductors for ADAS applications is totally different from mechanical transmissions. You have a new generation of IC components every 9-18 months, whereas you might be building the same transmission for five to seven years. Covid revealed that the entire automotive industry did not have the risk visibility across the value chain to manage the dynamics inherent in electronics.
The automotive industry suffered the most pronounced economic impact of an estimated $220 billion in lost sales. My view is that the auto sector will continue to suffer into 2023.
Q: The auto industry has made noises about developing closer relationships with semiconductor companies, perhaps bringing chip manufacturing in-house, which would be a significant shift in the business model. Do you see this happening?
RB: Yes, we do see that happening. Some of that was happening before Covid. Our view is that generally, it makes sense because high quality is critical for the automotive experience, whether it’s a passenger car or a truck. We know that the technology behind digital mobility services is defining the vehicle’s value today. The traditional aspects of horsepower and transmission are taken for granted.
But chip design is special. It’s different than building a transmission. You must invest in the right talent. Large organizations have the capacity to do that, and many of them were already doing so for specific elements of their next-generation platforms, especially electronic vehicle manufacturers. They are building ecosystem strategies around chip design and software with new partnerships involved in the platform decisions. That was already underway before Covid.
It’s central to the plan to de-risk the supply chain. But it’s a balancing act. You don’t want to vertically integrate or design in your own custom chipset because, relatively speaking, vehicle production volumes are 1,000 times smaller relative to consumer devices like smartphones. Automakers don’t want to vertically integrate because they are not going to win.
The compromise is to de-risk by multisourcing critical semiconductors and components. Use standard components where possible and have multiple suppliers or alternate form-fit-function parts and components designed in that offer more flexibility as market conditions change.
For example, at Tesla, if the part is critical to a subsystem being designed with manufacturing partners or tier-one suppliers, it will take more control, either of the chips they develop, or more importantly, of the negotiated long-term agreements with the semiconductor partners. Also, it may make sense to invest in the risk premium to protect that capacity. This is a novel suggestion as the automotive OEMs have never taken on that liability even though long-term purchase agreements in the more mature mobile, communications, data center companies are common practice.
The bigger story is how automotive OEMs will move from an arm’s length relationship to long-term agreements and risk-sharing relationships.
Q: What is the impact on high-tech supply chains from the ratcheting up of geopolitical tensions?
RB: The United States has just come off a series of saber-rattling, highly variable tariff threats with China. And Japan and South Korea have renegotiated a bilateral trade agreement. In addition, the USMCA free trade zone in Mexico has introduced new incentives for customers and local content and is ramping up automotive production capacity.
The escalating tensions between the U.S. and China are multi-dimensional. First, it was the tit-for-tat trade sanctions under the Trump administration, then a rise in tensions over US access to the China Sea, and the threat to Taiwan’s independence. We have moved into a strategic reset about where the sources of supply need to be and where companies want to invest. The tension has created a lot of uncertainty because there’s a near-term need for more flexibility, but we’re not sure what the new normal looks like in terms of investment decisions.
OEMs are looking at capital expenditures, such as bringing new fab capacity online or building new manufacturing capacity. Those decisions take three to four years just to get to runtime, much less to get to a positive ROI. On the one hand, we’ve seen a very healthy diversification of new foundry capital investment committed by multiple players that should diversify and improve the overall semiconductor market situation. But you still have a lot of highly sensitive geopolitical dynamics around future access to Taiwan and import restrictions for ASM technology into China. Meanwhile, China is ramping up its domestic infrastructure around chipset design and foundry and fab manufacturing technology.
It’s about access to the underlying technology for advanced semiconductor manufacturing. These issues will continue to play out. I found it interesting that Deutsche Bank has a high-tech index. They were looking at estimating what the total impact of decoupling would be: the answer is up to $5 trillion of additional cost and investment over multiple years to theoretically replicate all the deeply intertwined relationships. Practically, it’s just not possible to decouple fully.
We’ve got to watch the mounting geopolitical pressures very carefully. I think we will see more public-private discussions and partnerships starting to happen with the White House. We are involved in providing insights and feedback to different Congressional committees looking for outside advice to shape policy strategy in the future. And we are seeing similar conversations taking place in the EU and other places that will lead to public-private partnership alignment. Perhaps new forms of directed industrial policy for long-term incentives and capital commitments.