Skip to contentSkip to footer
People
On December 6, AlixPartners hosted an event for semiconductor executives in Silicon Valley alongside Vox Media. The event kicked off with a moderated discussion led by Markus Bolte, AlixPartners’ global co-leader of its semiconductor practice, on three crucial challenges semiconductor manufacturers must confront in 2025 and beyond. This was followed by a live Decoder podcast interview hosted by The Verge’s Deputy Editor Alex Heath in conversation with Arm CEO Rene Haas.
As the global semiconductor market continues to rapidly expand towards its trillion-dollar annual revenue ambition—by a projected 15% in 2025 alone, per IDC data—it’s imperative that manufacturers strategically plan to navigate headwinds brought on by geopolitics, supply and demand cycles, and new technologies like AI.
How will U.S. – China relations, among other geopolitical questions, affect semiconductor operations?
According to the 2025 AlixPartners Disruption Index (ADI), based on a survey of more than 3,200 senior global executives, 87% of companies across segments are holding onto more inventory than normal due to geopolitical conflicts. And as Donald Trump takes office for the second time, companies are watching how the economics of the U.S., China, and the rest of the world shift. These economies are inextricably linked to an extent that a separation of supply chains or technology would be incredibly difficult to architect. Yet rising geopolitical tension could lead to trade wars or new tariff programs, necessitating a rethink of global semiconductor operations. How might these dynamics impact the industry, and how will supply chains need to adjust in response?
One thing is for certain: China is one of the leading engines of innovation and will continue to challenge many players in the semiconductor world with rapid innovation cycles. Driven by hefty investments in equipment and new fabrication facilities, China’s production capacity is projected to increase by 40% over the next five years, according to TechInsights data. It will be critical for western companies to maintain their presence in China to stay competitive and attuned to that market—and 83% of ADI respondents plan to increase their investments in China over the next 12 months. How easy it is to do so will depend heavily on what changes with the new U.S. administration and on how China responds.
When will the AI hype cycle peak and where will it turn next?
We are early in the AI maturity cycle. Over the next few years, as AI expansion proliferates and the technology further embeds into all work functions, a key question for companies is how to leverage AI to improve employee productivity. Many worry AI will take their jobs, but we believe it will instead enhance workforce effectiveness by taking on repetitive tasks and allowing employees more bandwidth for high-value work, thus counteracting a potential shortage of qualified labor in the decades ahead. As companies continue to integrate AI into their organizations, it’s crucial they take an adaptive human-resource-management approach, with a focus on change management that helps employees see AI as a colleague rather than competition.
In the next five years, the focus of AI is expected to shift from training to inference, with emphasis on efficient usage of graphics processing units (GPUs) used to deploy AI models. But we are still seeing massive investments in training, which is hugely compute- and power-intensive. The level of compute needed for inference is actually much larger, and in the future, 80% of AI workloads may be inference-based. This leads to questions of whether data centers will be able to keep up with this demand—in a sustainable way, or at all.
AI integrated circuits (ICs) currently cost around $30,000 each—a high figure, though one that is bound to lessen significantly in the years to come per Moore’s Law (and increasing competition as startups challenge the incumbents). “We’re at the very beginning of the [AI] cycle,” said AMD CEO Lisa Su, “and the beginning of AI is bigger than [the beginning of the PC, or mobile phones, or the cloud].” The consensus of event participants was that it won’t be long before AI will move from hype to true value creation as more sophisticated business models and applications are developed. “It will come slowly and then suddenly” as one participant said.
Figuring out where AI is on its hype curve in each specific sub-segment of the market, and particularly learning how AI will impact semiconductor design and manufacturing, will determine how semiconductor manufacturers can best utilize the technology in the short and long term.
How do semiconductor players best manage market cycles?
The growth in semiconductor manufacturing has healthily recovered since its slowdown in 2022 and 2023, but balancing supply and demand forces is still paramount for industry players. Suppliers added $25 billion of “just-in-case” inventory after pandemic-induced demand shortages, and as a result, chip markets for several sectors are still saturated. A decline in demand, particularly in automotive, has led to difficulties for manufacturers and channel partners looking to offload inventory and prices and profits are dropping sharply.
What will drive the next upturn and when will it come? These are the critical questions those in the industry are trying to solve. The demand for chips with more compute power, more networking bandwidth, and more storage is only going to increase in the data center and cloud markets as more AI use cases are identified. Growth in edge- and device-based AI will also necessitate more compute power closer to the end users.
Electric vehicles (EVs) and software-defined vehicles (SDVs) are among the most prevalent end users, and we’re witnessing a shift in the relationship between the automotive original equipment manufacturers (OEMs) and semiconductor makers and distributors as the market participants struggle to discover their future playing field. China has an early lead due to its lower cost of production and much faster innovation cycles in developing new platforms and products. How U.S. and European companies adapt their business and operating models to compete with China in the EV market and the adjacent automotive semiconductor markets will be important to watch.
As different trends fuel different semiconductor market cycles, industry players will also aim to pass on more of the financial risk to end users—which could do away with cycles entirely.
Companies need to carefully observe these trends as they seek to maintain market competitiveness. Alex Heath’s interview with Rene Haas dives deeper into the semiconductor industry through a CEO’s perspective—listen or read here to learn more.
