Electronic Design Automation (EDA) is the bedrock of modern digital transformation, encompassing software-based tools used by all semiconductor companies to design cutting-edge chips and complex integrated circuits that power our technological landscape.
EDA has, accordingly, accompanied the growth of the semiconductor industry over the years. Looking forward, the growth pace of the main EDA players, including Synopsys, Cadence Design and Ansys, could hit an inflection point and materially exceed that of the semiconductor industry. While the AI boost to chip design activity will be a positive for the whole industry, AI will also provide a new monetization opportunity to software vendors in the form of AI-based software tools expected to accelerate the chip design process and ease engineering shortages.
The AI craze has opened a new growth era for semiconductors. AI use cases are indeed proliferating in most end markets, sparking a surge in the number of chip architectures and the multiplication of end market or application-specific AI semiconductors. The current AI chip market, estimated around $25 billion according to various sources, is then expected to exceed $100 billion by 2030. Importantly, the evolutions and advances in the AI chip segment have knock-on consequences on surrounding semiconductors such as storage, connectivity, sensing or power management chips. Overall, this new growth era, as we said above, should translate into a massive design activity from semiconductor companies, an obvious positive for EDA software solutions.
Amid increasing pressure on semiconductor companies to deliver on this market growth and produce efficient chips based on tailored designs, a shortage of engineers and architects is emerging as a significant constraint, underscoring the necessity for innovative approaches. Estimates project a design engineering gap of 15% to 30% by 2030, along with an anticipated 50% surge in chip design demand, highlighting the imperative of greater automation in semiconductor design. In this context, artificial intelligence has emerged as a pivotal solution to bridge this gap.
With the ability to produce a broad spectrum of human-like content, including intricate programming code, generative AI has demonstrated its potential across a wide range of applications, including the semiconductor design sector. AI-driven EDA is currently causing a paradigm shift in chip design and offers a host of tangible advantages that are poised to redefine the semiconductor industry in the years ahead.
Prominent EDA vendors have already bolstered their offering to provide AI-enabled software catering to chip design. And as these pioneering AI-based tools progressively integrate complex sets of chip conception processes, they are triggering profound changes, offering potential for faster, better, and cheaper semiconductors.
This potential stems from the substantial boost in engineering productivity facilitated by AI-driven EDA tools, with designers experiencing accelerated design cycles, improved design quality and superior performance in conceiving more compact and efficient chips that consume less power. This game-changing disruption means that tasks once requiring months of collaborative effort from an entire engineering team can now be accomplished within weeks by a single engineer.
Nvidia, for example, is using AI and machine learning to accelerate chip design (by up to 30 times), resulting in faster outcomes and qualitative advancements. Fabless chip designer ARM, which is currently in the process of going public again, offers another illustration: it recently reinforced its AI-enhanced design partnership with Synopsys to target intricate mobile chip designs, central to companies like Qualcomm, Apple, and Samsung.
As AI-driven EDA is boosting the productivity of semiconductor design by at least 20%, the manufacturing side is also seeing a significant performance boost. For example, Nvidia’s proprietary cuLitho software library (running on its GPUs) accelerates the production of photomasks used to “print” chips by a factor of 40x. AMD and others are also integrating AI in the fields of verification and testing – with reports showcasing reductions in silicon test costs exceeding 20%.
Given these productivity gains, it is obvious that the AI monetization opportunity is significant for Synopsys, Cadence and peers. Based on the initial AI pricing initiatives and feedback from other software makers, we believe that an AI-powered EPS upside of 20-30% is achievable in the next couple of years. This, combined with strong semiconductor design activity, suggests that EDA players are on track to deliver double digit top-line and earnings growth over coming years, with high visibility, making them highly attractive investments in the semiconductor space.