Does CADA promote European-made processors and accelerators?

Yes, as proposed, the Cloud and AI Development Act (CADA) explicitly promotes the development and deployment of European-made processors, accelerators, and

Summary Yes, as proposed, the Cloud and AI Development Act (CADA) explicitly promotes the development and deployment of European-made processors, accelerators, and quantum accelerators. Under Article 4(2)(b), the legislation mandates support for "AI-optimised servers and baseline software based on processors, accelerators and quantum accelerators designed and manufactured in the Union." This provision is a core component of Operational Objective 2, which aims to build secure, resilient, and performant open cloud computing stacks that support the Union's technological autonomy. By linking hardware manufacturing to software stack development, CADA seeks to reduce strategic dependencies on third-country providers and strengthen the European semiconductor and cloud supply chains, working in tandem with the Chips Act.

Detail

The proposed CADA addresses the EU's critical dependence on non-European cloud providers and the underlying hardware that powers them. A significant portion of the regulation's supply-side measures focuses on bridging the gap between advanced research, semiconductor manufacturing, and cloud infrastructure deployment. The primary mechanism for promoting European hardware is found in Title II (Research, Development and Deployment Activities for the Cloud and AI Ecosystem), specifically within the Cloud and AI Leadership Initiatives.

Article 4: Operational Objectives and Hardware Promotion

Article 4 of the proposal sets out the specific operational objectives of the Cloud and AI Leadership Initiatives. While the general objective (Article 3) focuses on promoting research and achieving large-scale capacity, Article 4 provides the actionable framework for achieving these goals through specific technological targets.

Operational Objective 2 is dedicated to "supporting the development and deployment of cloud computing stacks supporting the Union's technological autonomy." Within this objective, Article 4(2)(b) explicitly states that the initiatives shall:

"(b) develop AI-optimised servers and baseline software based on processors, accelerators and quantum accelerators designed and manufactured in the Union, alongside next-generation ultra-high density and long-term data storage;"

This provision is significant because it does not merely encourage the use of European hardware; it mandates the development of the entire stack, from the physical silicon to the baseline software that runs on it. By specifying "designed and manufactured in the Union," the proposal aims to create a closed-loop ecosystem where European intellectual property and industrial capacity are retained within the EU. This is intended to prevent the scenario where European software relies on proprietary, third-country hardware that may be subject to extraterritorial jurisdiction or supply chain disruptions. The inclusion of "quantum accelerators" further signals a forward-looking approach, ensuring that the EU's future compute infrastructure is not only sovereign today but also prepared for the next generation of computing paradigms.

Link to Grand Challenge 2: Cloud Stacks

The promotion of European processors and accelerators is further reinforced by Annex I of the proposal, which outlines the "Grand Challenges." Grand Challenge 2, titled "Cloud stacks," focuses on "Building end-to-end hardware and software cloud stacks, including AI tools, infrastructure, services and management layers to bridge the Union's critical capacity gaps."

The Annex specifies that this includes:

"Building AI servers powered by semiconductors and quantum technologies designed and manufactured in the Union for distributed and decentralised cloud and edge computing for AI."

This creates a direct legislative link between the high-level operational objectives in Article 4 and the strategic grand challenges in Annex I. The goal is to create "European open cloud stacks" that are not only technologically competitive but also sovereign. By integrating European-designed semiconductors into these stacks, the EU aims to ensure that the foundational layer of its cloud infrastructure is under European control, thereby enhancing security, resilience, and strategic autonomy. The text explicitly notes that "Pilot programmes could help demonstrate the capabilities of the European open cloud stacks in strategically important sectors," suggesting that the hardware-software integration will be tested in real-world environments before broader deployment.

Connection to the Chips Act

CADA is designed to work in synergy with the Chips Act (Regulation (EU) 2023/1781). Recital 14 of the CADA proposal explicitly mentions the need to "facilitate the development of AI-optimised servers and software including processors and accelerators manufactured and designed in the Union, such as those developed under Regulation (EU) 2026/XXX [Chips Act 2.0]."

While the Chips Act focuses on increasing the EU's share of global semiconductor manufacturing and securing the supply chain, CADA focuses on the application and integration of these chips into cloud and AI systems. The two instruments are complementary: the Chips Act ensures the physical production capacity exists, while CADA ensures there is a viable market and technical ecosystem for these chips by promoting their integration into sovereign cloud stacks. This holistic approach addresses both the supply-side (manufacturing) and demand-side (cloud adoption) aspects of technological sovereignty.

The proposal also notes in Recital 27 that the Cloud and AI Leadership Initiatives should build on "grand challenges" established in the Chips Act 2.0, "aimed at enabling semiconductor technologies underpinning AI, cloud computing, data centres and edge infrastructures." This confirms that CADA is not a standalone initiative but a critical demand-side lever that ensures the chips produced under the Chips Act have a dedicated, high-value application within the EU's digital infrastructure.

Strategic Autonomy and Security

The push for European processors and accelerators is not merely economic; it is also a security imperative. Recital 46 highlights the risks associated with dependence on third-country providers, including "vulnerabilities arising from the extraterritorial application of third-country laws" and "reduced control and oversight over personal and non-personal data and infrastructure." By promoting hardware designed and manufactured in the Union, CADA aims to mitigate these risks. European-made hardware is presumed to be less susceptible to foreign interference, backdoors, or sudden supply cuts, thereby enhancing the overall security posture of the EU's cloud and AI ecosystem.

Furthermore, Article 4(2)(a) supports the development of "advanced data centre technologies incorporating principles of energy efficiency and resource efficiency by design," while Article 4(2)(b) ensures these technologies are powered by Union-made silicon. This combination aims to create a sustainable, sovereign, and high-performance infrastructure that can support the growing demands of AI workloads without relying on external supply chains.

What this means for you

For CTOs, architects, and SMEs evaluating the practical impact of CADA, the promotion of European processors and accelerators presents both opportunities and strategic considerations:

Funding and Support Opportunities: As a provider or developer, you may be eligible for support under the Cloud and AI Leadership Initiatives if your projects align with Article 4(2)(b). This includes grants or subsidies for developing AI-optimised servers using European-designed chips. SMEs should monitor calls for proposals related to Grand Challenge 2 and Operational Objective 2, as the proposal explicitly aims to "foster the development of cloud computing stacks alternatives for strategic sectors."
Procurement Preferences: While CADA does not mandate that all public sector procurement must use European hardware, it encourages the use of sovereign cloud services. As the EU develops its Union assurance levels (Article 16), there may be indirect pressure to use hardware that meets high sovereignty criteria. European-designed processors may be viewed more favorably in risk assessments for sensitive public sector activities (Article 29), particularly where "operational autonomy" is a key requirement.
Supply Chain Resilience: Integrating European-designed processors and accelerators into your cloud stack can enhance your resilience against geopolitical shocks. By diversifying away from single-source, third-country dependencies, you can reduce the risk of supply chain disruptions. This aligns with the broader EU goal of creating a "resilient, high-performance EU cloud and AI ecosystem."
Technical Compatibility: The proposal emphasizes the development of "baseline software" alongside hardware. This means that adopting European processors may require adapting your software stack to work with new architectures and instruction sets. Early engagement with European chip manufacturers and participation in open-source software foundations (supported under Article 4(2)(d)) can help ensure compatibility and optimize performance. The text notes that the initiatives shall "foster the creation of open-source software foundations supporting open-source components," which could be a key avenue for software adaptation.

Common misconceptions

Misconception: CADA bans non-European hardware.
- Reality: CADA does not prohibit the use of non-European processors or accelerators. Instead, it provides financial and regulatory incentives to develop and deploy European alternatives. The goal is to increase market share and availability of European options, not to exclude others through outright bans. The focus is on "supporting the development" and "fostering" alternatives, not on prohibition.
Misconception: Only large tech giants can benefit.
- Reality: The proposal explicitly includes measures to support SMEs and start-ups. For example, Article 33 encourages monitoring and reporting on procurement of innovation, with a target that at least 25% of relevant cloud and AI procurement innovation procedures be awarded to SMEs. Additionally, the Cloud and AI Leadership Initiatives include operational objectives to support the uptake of AI technologies by SMEs (Article 4(2)(h)), ensuring that the benefits of European hardware development are accessible to smaller players.
Misconception: CADA replaces the Chips Act.
- Reality: CADA complements the Chips Act. The Chips Act focuses on manufacturing capacity and supply chain security, while CADA focuses on the integration of these chips into cloud and AI services. They are part of a broader EU strategy to strengthen the entire digital value chain, from silicon to software. Recital 14 explicitly links the two, noting that CADA supports technologies "such as those developed under Regulation (EU) 2026/XXX [Chips Act 2.0]."
Misconception: "Designed and manufactured in the Union" applies only to the final server.
- Reality: The text specifies "processors, accelerators and quantum accelerators designed and manufactured in the Union." This implies a requirement for the core silicon components themselves to be of Union origin, not just the final assembly of the server. The focus is on the foundational hardware layer.

This is general information about a draft EU regulation, not legal advice.