Back

What is to come in the AI semiconductor market?

The global semiconductor market has rallied by more than 50% in little more than 4 months. This substantial performance comes on the back of strong and global AI infrastructure spending. While the market is fueled by general hype around AI, there are some valuable developments taking place that justify a $7 trillion investment plan by OpenAI’s CEO Sam Altman.

No doubt, the underlying business fundamentals in the semiconductor industry are positive. Yet, many investors started to argue that the core AI companies are overvalued and that the rally will fizzle out in the case the announced spending boom does not materialize. 

To start with, let’s look at the basics of the sector. Most of the leading semiconductor companies, including AI companies, are cash-rich and generate strong free cash flows, hence there is no cash shortage to feed the required Capex. However, in this stage of the AI game, the availability of cash is only part of the equation. The more important part is the creation of intellectual property (IP) and how the latter is interconnected with processes, databases, and ledgers. Given that the field of AI is relatively new, the engineering capacities can be generated by only a few brains, and hence only select companies will be able to engage in the process and capture the addressable market.

In the financial year 2024, the semiconductor sector will generate revenues expected at 700’000 million, the investment plan of Sam Altman is 10x this figure. As of now, the most advanced chip technologies are in the field of 5nm. Assuming that 10% of the spending goes into semiconductor equipment, we may potentially see an additional capacity of 2.8mn per month. According to UBS research, this represents twice TSMC’s current installed capacity, which took many years to build.

Semiconductor equipment companies Given that any such large-scale investment plan will take significant time to alter the present market capacity, one can expect little to no change in the industry set-up. Therefore, names such as Nvidia, TSMC, Broadcom, ASML, and AMD are expected to expand the lead.

The list of companies with a catch-up potential includes Micron, edge-AI beneficiaries like Qualcomm and TSMC, and chip design software companies like Cadence and Synopsis.

The largest semiconductor equipment companies that will benefit from the secular growth trend are: ASML, KLA, AMAT, Tokyo Electron, and LAM.