In the United States and Europe, the goal for both the chipmakers and policymakers is likely to make their domestic industrial capacity appropriately more self-sufficient while recognising that self-sufficiency may be unattainable. As Ursula von der Leyen, president of the European Commission, said in her remarks introducing the EU Chips Act: “It should be clear that no country—and even no continent—can be entirely self-sufficient.”

After all, “chips” is not a monolithic category: Many kinds of chips (memory, logic, mixed signal, power semis, etc.) are manufactured for many different end markets. Each type of chip may require different wafer sizes and process technologies as well as different materials, facilities, equipment and design tools, radiation tolerance and so on. There are two dominant but different models: the fabless foundry outsourced assembly and test (OSAT) ecosystem; and the integrated device manufacturing model, which usually combines all three roles.

Deciding what mix of onshoring, nearshoring and friendshoring (while almost certainly still relying on offshoring as well) may be best for each manufacturer or country/region will make 2023 an interesting year. These decisions could resonate for years to come: New or expanded facilities started in 2023 will likely still be in operation in 2030 and beyond, and so will their supply chain linkages.

Strategic questions to consider, by region or country:

• Will US labor costs and the focus on higher margin aspects of the value chain keep assembly and testing (AT) offshore? Is nearshoring in Mexico and Latin America—especially for some specialty chip assembly, test, and module assembly—a viable option? What should be the fab versus foundry mix to help support the US-based fabless companies that have strong design competencies? 
• Should Europe rely more heavily on friendshoring? How much advanced-node chipmaking does Europe really need given that the auto and other industries in Europe rely much more on trailing nodes? In terms of AT, can Europe turn to Eastern Europe? Given the other big gap is in chip design, in which Europe trails badly at present, what should it do to address this issue? 
• Meanwhile, Taiwan, China, Japan, and South Korea, already dominant in many parts of the semiconductor ecosystem, are fighting hard to keep their preeminent positions and market share (or even grow them). For these regions, what factors—including threats and challenges—might adversely affect their positions in the near term?
• Many other countries—India, Canada, Saudi Arabia, others—are making similar choices although usually at a smaller scale. What key parts of the semiconductor supply chain do they really want in their country? What can they afford—and what is realistic?

The US and Europe chip industries are looking at diversifying not just fabs but all parts of the semiconductor supply chain, including AT. They would move chipmaking out of the traditional strongholds in the Asia/Pacific region into North America and Europe.

Replicating the capabilities of Asian manufacturing locations won’t be easy. Until the supply chain, pandemic and trade issues surfaced, Asia had secured supply of raw and manufactured materials to make hundreds of components. Replicating and building that model in multiple geographically unconcentrated new locations would likely take years or even decades. Nonetheless, starting in 2023, the chip companies need to plan and prepare for potential risks involved in diversification.

One of the crucial starting steps is expected to be identifying new sourcing and trade routes. The ongoing war in Ukraine and US-China technology restrictions introduce additional complications and risks and will likely require new air corridors, sea routes and export-control compliance. These new pathways are likely to increase freight, transportation, and logistics costs and, in turn, pricing of the end products—which are already elevated due to higher energy costs.

With supply chains most likely to fan out to dozens of new locations within the next two to three years, semiconductor companies may need to invest in advanced software and analytics in 2023 to track their products and solutions at any given point in time across the supply chain and facilitate timely decisions.

Strategic questions to consider:

• While diversifying the manufacturing base and supply chain, what factors should be considered related to building their own facilities, buying existing ones, partnering with other companies or peers, or friendshoring? 
• When commencing operations in new regions and fostering new alliances and partnerships, what types of intellectual property, technological, trade restriction and cyber risks could emerge?
• Given the current downturn and falling profits, how can companies preserve cash while expanding into new geographic regions, pursuing strategic M&A, investing in new joint ventures or alliances and divesting select non-core assets?

The global semiconductor industry is anticipated to grow to US$1 trillion in revenues by 2030, doubling in this decade. This growth is expected to require investment in high-end advanced wafer manufacturing materials, equipment, and services—but also in the back-end AT solutions and services.

Currently, more than 90% of the AT base is in Asia, either in foundries or OSAT providers. With around US$100 billion pledged to support localisation of semiconductor capacity, several companies in the industry are adding new AT capacity in Europe, North America and Southeast Asia. And that could exacerbate the existing challenge of disparate and siloed supply chain tech and systems, rendering it even more difficult to predict potential supply chain disruptions. This is where integrated data platforms, next-generation ERP, planning and supplier collaboration systems, along with AI and cognitive technologies, are expected to make OSAT processes more efficient and help sense and preemptively plan for future supply chain shocks.

Beyond OSATs, data analytics platforms that are integrated into ERP, planning, and procurement systems can help semiconductor companies predict events that could disrupt the supply chain: unexpected weather events, transport bottlenecks, logistics challenges that require rerouting shipments and labour-related issues. Sharing real-time data and intelligence across the ecosystem—encompassing the equipment and tool suppliers, chip design companies, wafer fab facilities, AT facilities, distributors, OEM and ODM customers and more—is likely essential to build a digitally connected supply chain.

To realise maximum benefit from AI tech and data-driven solutions, data quality matters. So in 2023, semiconductor companies need to modernise their ERP systems and integrate diverse data sources, such as customer data, manufacturing data, and financial and operational data. And they should consider investing in data management and data analytics modernisation solutions, establishing the required data governance and fixing data quality issues.

Strategic questions to consider:

• Given the current economic and industry trends (rising interest rates, falling demand in some important chip sectors, high inventories), how can companies best revamp and accelerate digital transformation in the face of the adverse trends? 
• When deploying technology and AI tools to digitise supply chain management, what entities and variables should be considered as part of the advanced and predictive analytics data-driven decision model? 
• As localisation accelerates through the next years, how much of a negative impact will there be on the supply chain and data modernisation efforts? How do companies effectively interconnect new and existing entities (e.g., a fab or site) into an increasingly complex global network?
• How can companies deploy resources to modernise and integrate their IT systems when budgets are getting tighter? And how can they balance near-term IT implementations with strategic long-term investments in next-generation digital technologies?

Semiconductor talent was in short supply in 2022—and the shortage is expected to get even worse in 2023 for parts of the industry and be a challenge for the rest of the decade. The race to localise semiconductor manufacturing is intensifying globally and is exacerbating the chip talent and skill shortage issues. Deloitte predicts that the semiconductor workforce—estimated at more than 2 million direct semiconductor employees worldwide in 2021—will need to grow by more than 1 million additional skilled workers by 2030, adding roughly more than 100,000 workers annually.

In 2023, chip companies should consider expediting hiring people with diverse skills for building and automating their manufacturing facilities and to design chips and tools. Their talent challenges are compounded by the urgent need to build large-scale fab facilities in multiple regions. Therefore, they need to accelerate hiring for a range of skills: electricians, pipefitters and welders; technical engineers, maintenance personnel and smart factory automation specialists; and graduate electrical engineers to design chips and the tools and manufacturing processes that make the chips.

Besides skill development, the chip industry may need to join hands with the various local governments to enhance fungibility and mobility of skills between regions, secure support in the form of favorable talent immigration policies and seek assistance for local recruitment and skills-based training. Further, after years of M&A activity and consolidation, many companies may need to find a better way to integrate the various talent workforces.

Strategic questions to consider:

• What specific efforts and programs are required to identify, access, and develop future skills across the engineering and manufacturing workforce and drive superior performance and value? And how can semiconductor companies plan and prioritise hiring in key roles and expand their workforces while dealing with hiring freezes and limited budgets?
• How can semiconductor companies redesign the way humans and technology interact to develop solutions and deliver services and value for customers? Does that need to be about advanced tools such as AI, automation, robots and co-bots, or can the industry benefit from simpler solutions?
• What should an organisation’s talent design model look like to ensure that it not only addresses the immediate talent-related issues but also allows the organisation to capitalise on new or greenfield areas that offer long-term growth opportunities?

The global semiconductor industry has experienced the impact of climate-driven adverse events in 2021 and 2022. Extreme heat and drought conditions affected power supply in China’s Sichuan province, leading to factory closures at several electronics component makers. Drought in Taiwan caused water restrictions to some chipmaking plants. Intense winter storms triggered power outages in Texas, affecting three major semiconductor plants’ operations.

But the chip industry is not just facing climate issues; it is likely contributing to climate change. The production process for every new generation of chips uses more energy, water and greenhouse gases—especially process gases with high global warming potential (GWP) that are difficult to mitigate—than the generation before. And by 2030, the information and communications technology industry is likely to account for 20% of global electricity demand. As a result, investors, customers, board members and regulators are requiring more transparent and comprehensive disclosures on greenhouse gas emissions, environmental risks and mitigation actions.

In 2023, the chip industry can form strategic alliances with all parts of the supply chain and work more cohesively to explore and develop new technologies and methods to help accelerate decarbonisation efforts—although there is some skepticism that the industry can achieve net-zero. Besides reducing carbon emissions from their own operations, the chip industry has the potential to reduce others’ carbon footprints. Compound power semiconductors, such as silicon carbide and gallium nitride, can help enhance energy efficiency and improve the green carbon footprint of sectors such as automotive, industrial automation, rail transportation and renewable installations.

Strategic questions to consider:

• What broader sustainability and net-zero strategy will likely be necessary as companies build out new fabs at an unprecedented pace and in new regions?
• When defining sustainability and net-zero goals, what specific parts of the upstream and downstream supply chain should be prioritised over the next two to three years, based on what would be feasible?
• Will governments, regulators, stock exchanges, or shareholders demand semiconductor companies move significantly faster on sustainability during 2023?
• Are irreplaceable process gases (which have high GWP but are essential and difficult to mitigate) the biggest problem for the industry?