Hardware·Americas

Siemens and Synopsys Shift Semiconductor Approach at Sub-2nm Scale

Global AI Watch · James Harrington··5 min read
Siemens and Synopsys Shift Semiconductor Approach at Sub-2nm Scale
Point de vue éditorial

The shift to sacrificial pads and hybrid bonding at sub-2nm nodes marks the third major process evolution since 2024.

What Changed

The semiconductor industry is witnessing a significant shift in manufacturing processes at sub-2nm nodes, driven by Siemens EDA and Synopsys. Siemens, under the leadership of Lee Harrison, is transitioning from probing chiplets to using sacrificial pads to prevent damage during the manufacturing process. Simultaneously, Synopsys, led by Rob Kruger, is adopting hybrid bonding technology, which offers a more robust solution for bonding ultra-thin dies and wafers. This shift is primarily due to the increased sensitivity and reduced margin for error in these advanced nodes, where even a single atom can alter chip behavior.

The adoption of these technologies is not entirely new but marks a critical evolution in semiconductor practices. At sub-2nm nodes, traditional manufacturing processes used at larger nodes fail to meet the precision requirements needed. Consequently, companies are moving towards redundant server clusters in AI data centers to mitigate potential manufacturing defects. This approach, reminiscent of mainframe strategies, allows for hot-swapping of resources, thereby reducing downtime and enhancing reliability.

Strategic Implications

The strategic implications of this shift are profound for the semiconductor industry. By adopting sacrificial pads and hybrid bonding, companies like Siemens and Synopsys are positioning themselves at the forefront of semiconductor innovation. This transition enhances their competitive edge by addressing the challenges posed by thinner and more fragile chip structures. Additionally, the move towards redundancy in data centers suggests a shift in how AI infrastructure is designed, focusing on reliability and fault tolerance.

This evolution also signals a broader trend towards national semiconductor autonomy. As companies develop more sophisticated manufacturing techniques, countries may push for localized production to secure their supply chains and reduce dependency on foreign technology. This shift could lead to increased investment in domestic semiconductor capabilities, particularly in regions aiming to establish themselves as leaders in advanced node manufacturing.

What Happens Next

Looking ahead, we can expect further advancements in semiconductor manufacturing technologies. By mid-2027, it is likely that more companies will adopt similar approaches to Siemens and Synopsys, focusing on hybrid bonding and redundancy in data centers. This trend will likely spur regulatory bodies to update standards and guidelines to accommodate these new technologies, ensuring that safety and performance benchmarks are met.

Moreover, as the demand for AI continues to grow, the need for more efficient and reliable semiconductor solutions will drive further innovation. Companies may also explore partnerships and collaborations to share technological advancements and reduce costs associated with manufacturing at such small scales.

Second-Order Effects

The adoption of these new processes will have several second-order effects. For one, the supply chain for semiconductor materials is likely to see increased demand for specialized materials used in hybrid bonding and sacrificial pads. This demand could lead to the development of new suppliers and the expansion of existing ones.

Additionally, the focus on redundancy in AI data centers could influence other sectors reliant on high-performance computing. Industries such as finance, healthcare, and telecommunications may adopt similar strategies to enhance the reliability of their systems, leading to broader changes in how data centers are constructed and managed.

Expert Perspective

From an expert viewpoint, this development signifies a critical juncture in the pursuit of semiconductor sovereignty. As companies like Siemens and Synopsys innovate at the sub-2nm scale, they contribute to a broader strategy of technological independence. This shift not only enhances their market position but also aligns with national interests in securing advanced technology capabilities. By fostering domestic innovation, countries can better navigate the geopolitical landscape and reduce vulnerabilities in their technology supply chains.

Free Daily Briefing

Top AI intelligence stories delivered each morning.

Subscribe Free →

Explore Trackers