NVidia soaring, Intel tanking - does that create opportunity?

I am willing to make a long term bet on Intel's capabilities

NVidia has emerged as the AI darling of the American investment community, coining massive margins by providing key chips to major AI leaders like Microsoft, Amazon, Oracle, Meta, and all the players in the race to develop large data centers to support their large language models and gain the upper hand in this rapidly emerging field. The big winner has been NVidia with its key supplier Taiwan Semiconductor turning in stellar results in parallel.

What about poor old Intel? It’s process technology has lagged TSMC, its proprietary chips are fading as conventional personal computers and servers have experienced soft demand, and its 115,000 person workforce seems bloated. CEO Pat Gelsinger has his hands full and is turning to Morgan Stanley and perhaps other advisors for a “strategic review” which is street-speak for separating its foundry business from its own chip business. The foundry side of Intel has been a laggard since Intel historically did little custom fabrication for others and companies like Apple and NVidia using ARM Holdings designed chips turned to TSMC to have their chips manufactured.

The outcomes show how important strategy can be in determining results. NVidia has a market capitalization approaching $3 trillion, TSMC is at about $1.2 trillion and once leading Intel has a market value of about $94 billion.

But there is more than a little hope for Intel. As NVidia advances its chips it is possible they will seek Intel as the source for its successor to Blackwell, while Blackwell is using TSMC’s now-dated 5nm process owing to some issues in TSMC’s efforts to make its 3nm process work well enough to keep up with demand for 3nm chips from Apple, who source their advanced chips from TSMC but TSMC is reporting some yield issues.

Intel’s bet is that it can bring its 2nm process to commercial stability by early 2025 and that is process will surpass the TSMC 2nm process. Microsoft has already chosen Intel for an undisclosed advanced chip, reportedly. One feature that distinguishes Intel is called “backside power delivery” which can improve the efficiency of their chips and may give them an edge, at least for the short term, over TSMC.

Yesterday’s market sell-off saw NVidai shed some $279 billion in market capitalization, the largest one-day drop for any stock in U.S. stock market history. Investors are sniffing out a reality that nosebleed prices for chips that are in demand attracts both competitors who will invest to build competing devices, prompts customers to seek alternatives, and triggers a DOJ review of monopolistic practices or similar acts to curb market forces. All are taking place.

Given their relative valuations and the market’s enthusiasm for NVidia and TSMC which is already embedded in their share prices (for good reason) the outlier is Intel and the potential gains if Intel can get its act together can be enormous. I missed the boat on NVidia and TSMC, but I think I will open a small position in Intel and wait and see. Sometimes, betting on the laggard works out in areas where the technical skills of the competing companies are comparable and the past differences reflect strategic choices rather than ability to execute. We will see.