I) Shortages, geopolitical tensions, concentration of resources: microelectronics under pressure

Over the past five years, the microelectronics sector has faced many disruptions, highlighting new vulnerabilities.

In our current societies, electronic devices, IT systems, and smartphones have become a must-have, driving a steady increase in demand for the electronic chips that power our favorite devices. During the Covid-19 crisis, this already rising demand skyrocketed due to the widespread adoption of remote work and digital collaboration, all while supply chains were disrupted (with longer transit times for instance), triggering speculation on component stocks. The combination of these factors led to an unprecedented semiconductor shortage, both in scale and duration (2020–2023). Notably, wafer prices for 300mm increased by 17% in 2022[1], with peak demand rising 26% for 28–40nm nodes[2].

Recent geopolitical tensions have also impacted the sector. Since 2022, the war in Ukraine has raised medium-to-long-term concerns about supplies of key raw materials for chip manufacturing: 90% of the ultra-pure neon gas used in lithography comes from Russia’s steel industry and is refined in Ukraine, and Russia supplies over one-third of the world’s palladium.[3] 2024 also saw the escalation of export restrictions from the US and China. Indeed, the US introduced major restrictions over advanced technologies with potential military applications, including advanced technological nodes and AI accelaration chips.[4] China retaliated with restrictions on gallium and germanium exports, which are critical materials for semiconductor manufacturing (of which China accounted for 59.2% and 98.8% of the global production in 2024, respectively).[5] Realization of these vulnerabilities in the semiconductor supply chain prompted the EU to inject 43 billion euros of investment by 2030 through the Chips Act to reduce these dependencies.

In short, recent events have exposed vulnerabilities in the semiconductor industry. Given the fragmented nature of the value chain, with high levels of country specialization coupled with a sometimes extreme concentration of resources and know-how (e.g. almost 75% of DRAM memory chips are manufactured in South Korea[6]), these disruptions are likely to recur. In such uncertain times, custom solutions are more relevant than ever to regain independence and control over the value chain. Beyond that, custom integrated circuits offer clear advantages in terms of performance, unit cost, differentiation, and security.

II) What is a custom integrated circuit?

Let’s start by defining what a custom integrated circuit — or ASIC (Application-Specific Integrated Circuit) is. Unlike off-the-shelf integrated circuits, which are designed for general-purpose use, ASICs are built for a specific application. They precisely match the expressed needs, with optimized performance.

An ASIC stands out from generic solutions in four key aspects:

	ASIC	FPGA	SoC
NRE Costs	1-5M$	50-500k$	2-10M$
Time-to-Market	18-24 months	3-6 months	12-18 months
Power Consumption	0.1-1W	5-10W	0.5-5W
Flexibility	Fixed	Reprogrammable	Partial

III) The benefits of custom integrated circuits

Custom ASICs offer many advantages across strategic areas:

• Performance optimization

Because ASICs are designed for specific tasks, engineers can fine-tune their performance for these functions. Removing unnecessary features also significantly reduces energy consumption (see, for example, the MASSAR IP).
Full-custom architectures, where each transistor is sized for its specific application, yield typical gains such as:

– 15-40% reduced latency via hardware pipelining

– 3-5x increased bandwidth through custom buses

– 2-4x improved energy efficiency thanks to adaptive voltage scaling

• Surface area optimization

Designing a custom integrated circuit also means optimizing its layout and architecture to reduce its final surface area. A smaller surface area reduces costs by lowering the materials bill. Reducing the surface area can also enhance the performance of the finished product, for example by integrating several custom ICs where it would have been possible to fit just one off-the-shelf IC. Finally, having an IC that takes up less space can also be critical in certain applications, such as wearable products[7].

• Lower marginal costs (no recurring costs)

Although developing a custom ASIC requires significant upfront investment linked to development costs, in the long term this investment allows freedom from recurring patent costs, thus increasing the value captured. Additionally, when you develop your own custom IC, you address the foundry directly, reducing manufacturing costs. This drop in recurring costs makes custom ICs particularly attractive for high-volume production.

• No dependance on third-party patents

Developing your own custom integrated circuit, or commissioning a Design House to do so, means obtaining intellectual property rights for the circuit. In this regard, choosing a custom IC eliminates the risks associated with dependence on third-party patents.

• Strategic Differentiation

Using custom integrated circuits is a strategic differentiation measure, and a means of gaining a long-term competitive advantage. It is a unique, proprietary solution to which no competitor will have access

• Enhanced security

In a world where cyberattacks are increasingly frequent and sophisticated — often powered by AI — cybersecurity is becoming a critical concern. Custom ICs help to meet this challenge by integrating cutting-edge hardware security mechanisms:

– PUFs (Physical Unclonable Functions) for unique authentication

– TEEs (Trusted Execution Environments) isolating critical processes

– Anti-EMA (ElectroMagnetic Analysis) countermeasures with differential shielding

These mechanisms secure the circuit at hardware level. Furthermore, developing an ASIC means that potential hackers are faced with a black box: the ASIC owner is the only one to know its contents, which enables him to protect his technology and know-how.

IV) When are custom integrated circuits most relevant?

Custom integrated circuits are obviously an ideal solution for addressing complex design and performance challenges. While customization offers virtually unlimited possibilities, ASICs are particularly recommended for projects with large production volumes, where ROI is maximized. Traditionally, companies operating in rapidly evolving markets were discouraged from using custom ASICs due to the length of development cycles that can take 12 to 24 months from design kickoff to manufacturing. However, FPGA to ASIC fast prototyping services now make it possible to significantly reduce development cycle times, opening up new prospects for custom ICs.

---

[1] https://www.eetimes.eu/silicon-wafer-shipments-up-3-9-in-2022-says-semi/

[2]  https://news.metal.com/newscontent/101609825/the-main-reason-is-that-the-price-of-silicon-wafer-increases-by-up-to-15-capacity-or-is-tight-to-accelerate-the-expansion-of-contract-factories-in-2023

[3] https://www.euractiv.fr/section/strategie-industrielle/news/la-guerre-en-ukraine-pourrait-impacter-la-production-de-semi-conducteurs/

[4] 2025 semiconductor industry outlook | Deloitte Insights

[5] China bans export of critical minerals to US as trade tensions escalate | Reuters

[6] 2025 semiconductor industry outlook | Deloitte Insights

[7] https://www.infineon.com/dgdl/Infineon-Benefits_Challenges_Custom_ASIC_Development_Bodo_072023-Article-v01_00-EN.pdf?fileId=8ac78c8c8929aa4d018977bfd95a2711