Which Semiconductor Design Stocks Could Benefit from the AI Revolution?

Behind every powerful artificial intelligence (AI) application is chip that was carefully designed using bleeding-edge tools and software. Here are some alternative stocks supporting the broader semiconductor ecosystem that may not be on investors’ radars.

• Synopsys and Cadence Design Systems have a combined 62% share of the global electronic design automation software market.
• Chip-equipment maker Lam Research sees long-term benefit from AI server demand.
• How to invest in the semiconductor ecosystem: the iShares Semiconductor ETF is up 24% in the past six months.

Generative AI may be garnering a lot of attention, but the semiconductor theme is about more than the chips that power AI applications. Chip production involves a number of stages, including design, testing and verification.

Virtually every chip in the world is designed with the help of software tools. This is because modern semiconductors have billions of tiny on–off switches, known as transistors, and their location on a chip will have a significant impact on performance and cost. Software from companies including Synopsys [SNPS] and Cadence Design Systems [CDNS] ensure the transistors are arranged as precisely as possible.

Synopsys and Cadence accounted for 32% and 30% of the global electronic design automation (EDA) software market in 2021, according to TrendForce research published last year. Their nearest competitor was Siemens-owned [SIE.DE] Siemens EDA, previously Mentor Graphics, which had a 13% market share.

The Synopsys and Cadence share prices have rallied this year, fuelled by the generative AI boom, which, along with government drives to shore up their domestic chip supply chains, have also given semiconductor equipment stocks, namely Applied Materials [AMAT] and Lam Research [LRCX], a major boost. The two toolmakers count both Intel [INTC] and Taiwan Semiconductor Manufacturing Company [TSM] among their customers.

The Stocks Sustaining the Semiconductor Ecosystem

Back in April, Lam Research published research concluding that, for optimal semiconductor design, “humans excel in the early stages of process development while algorithms are more cost-efficient near the tight tolerances of the target”.

Applied Materials launched a new range of tools in June that are able to handle more sophisticated chips. Heterogeneous integration, whereby multiple chips are combined into one package, will play a key role in addressing industry challenges as the number of AI applications, and demand for high-performance computing, continue to rise.

In February, the company also started selling technology that enables chipmakers to create high-performance transistors for AI tasks, but which involves fewer extreme ultraviolet (EUV) lithography steps, lowering the cost in the process. The machines could challenge ASML’s [ASML] dominance in the EUV lithography market.

In March, Synopsys rolled out new AI tools that enable better results during the process of chip design. Chipmakers can use the tools to detect defects in their designs, test sample chips from manufacturing partners and, once production is underway, ramp up the number of defect-free chips coming off the production line.

In March, Synopsys rolled out new AI tools that enable better results during the process of chip design. Chipmakers can use the tools to detect defects in their designs, test sample chips from manufacturing partners and, once production is underway, ramp up the number of defect-free chips coming off the production line.

Most recently, at the end of August Cadence announced a partnership with Arm to accelerate designs on the Arm Neoverse V2 platform. The collaboration has enabled Cadence to fine-tune the AI-powered processes for Neoverse V2 and deliver compatible 5nm and 3nm ‘rapid adoption kits’ to its customers.

AI Applications to Continue Driving Sales for Rest of 2023

In the long term, Cadence and Synopsys hope to continue benefitting from the AI trend by offering chipmakers software powered by AI, and by leveraging AI to enhance their own operational capabilities.

“This is going to take some time, which is okay. It is a natural process of trying and deploying,” said Cadence President and CEO Anirudh Devgan on the second-quarter (Q2) 2023 earnings call at the end of July. The company hiked its full-year revenue forecast on strong AI sales.

Synopsys Chairman and CEO Aart de Geus explained on the company’s Q3 2023 earnings call last month that it would be monetising AI “by pervasively embedding our pioneering AI across our full EDA stack” and “through AI-driven efficiency transformations, as we optimise and automate our own internal workflows”.

Synopsis revenue and operating margin forecasts for Q4 were raised, thanks in part to “robust design activity by our customers who continue to invest through semiconductor cycles”, said Chief Financial Officer Shelagh Glaser on the call.

Lam Research Sees Long-term Benefit from AI Server Demand

Lam Research is expecting a strong second half of the year, on the back of an uptick in “domestic China-related spending”. Revenues from China contributed 26% to total revenues for the quarter ending in June, and the company sees that trend continuing.

President and CEO Tim Archer also believes the company is only in the initial stage of benefitting from the AI trend. As chipmakers ramp up investments in factories to boost capacity and meet future demand for AI chips, the company thinks its tools will be critical over the next years.

“Advanced AI servers have significantly higher leading-edge logic, memory and storage content versus traditional servers, and every incremental 1% penetration of AI servers and data centres is expected to drive $1bn–$1.5bn of additional investment,” said Archer on the earnings call in July.

Looking Beyond the AI Hype

From an investment perspective, selecting the picks and shovels of the broader semiconductor theme and the infrastructure behind the ecosystem could be a wise decision, because “innovation doesn’t guarantee perpetual success”, argued Jeremy Schwartz, Global Chief Investment Officer at WisdomTree, in a blog post published last month.

Speaking about Nvidia [NVDA] specifically, Schwartz wrote that “the market, in general, is not in favour of a single supplier of AI computational power”. It’s important that investors “stay grounded and look beyond the present hype”, he added.

How to Invest in the Semiconductor Ecosystem

ETFs, or exchange-traded funds, offer an economical and diversified way to invest in a variety of stocks within a particular theme.

Funds in Focus: the iShares Semiconductor ETF

The iShares Semiconductor ETF [SOXX] has Applied Materials and Lam Research in its top 10 holdings as of 1 September. The portfolio has allocated 79.6% to semiconductors and 20.2% to semiconductor equipment. The fund is up 46% in the past year and up 23.9% in the past six months.

The Columbia Seligman Semiconductor and Technology ETF [SEMI] is another fund that has both Applied Materials and Lam Research in its top 10 holdings, as of 31 July. The portfolio is weighted 96.2% in favour of the information technology (IT) sector, with industrials and communication services accounting for the rest. The fund is up 30.1% in the past year and up 15.1% in the past six months.

The Global X Semiconductor ETF [SEMI.AX], which also holds Applied Materials and Lam Research in its top 10 (as of 6 September), is weighted entirely in favour of the IT sector. The fund is up 55% in the past year and up 29.9% in the past six months.