Former Intel CEO Pat Gelsinger bought NVIDIA stock dip amid DeepSeek sell-off

Investing.com -- Pat Gelsinger, the former CEO of Intel (NASDAQ:INTC), has purchased NVIDIA (NASDAQ:NVDA) stock during its recent dip, as he shared in a LinkedIn post discussing the ongoing DeepSeek situation. Gelsinger’s decision to buy comes as he offers his perspective on the market’s reaction to the DeepSeek development, suggesting that there are three key lessons to be learned from the past five decades of computing.

Gelsinger first pointed out that computing obeys the ’gas law’, meaning it fills the available space as defined by available resources such as capital, power, and thermal budgets. He used examples from the past, such as CMOS, PCs, multicore, virtualization, and mobile, to illustrate how making compute resources widely available at lower price points can drastically expand the market. He expressed his belief that AI will be integrated into everything in the future, and that the current cost of AI is too high to realize this potential. He stated that he is happy to benefit from lower prices by buying NVIDIA and AI stock.

His second point focused on the role of constraints in engineering. Gelsinger noted that the DeepSeek team faced numerous constraints but managed to deliver a world-class solution at 10-50X lower costs. He mentioned that export laws limited the resources available to the team, forcing them to be creative. Gelsinger shared an anecdote from his past, when he interviewed renowned computer scientist Donald Knuth, who explained how he did his best work when resources were most limited and schedules most demanding.

Lastly, Gelsinger emphasized the importance of openness in the field of AI. He expressed disappointment at the increasing closure of foundational model research in recent years, stating his alignment with Elon Musk’s stance that AI research needs more openness. Gelsinger stressed the need for transparency in training datasets, algorithms, and introspection on correctness, ethics, and implications. He referenced the success of open models in Linux, Gcc, USB, Wifi, and other examples, arguing that AI is too important to allow a closed ecosystem to dominate.

Gelsinger concluded by praising DeepSeek as an "incredible piece of engineering" that will prompt greater adoption of AI and help reshape the industry’s view of open innovation.

This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.