Alphabet wants to take over the position of the world’s largest company by market capitalization, a position held by Nvidia in recent months, which it will challenge in its artificial intelligence (AI) teams. Meta Platforms is in talks to invest billions of dollars in using Google’s parent company’s AI-focused chips.
In this scenario, Alphabet shares rose more than 3% this Tuesday in the first hours of trading, breaking again their all-time high above $328, giving the Google parent company a capitalization approaching four billion euros. At the beginning of the year, securities accumulate at revaluation of approximately 70%.
Conversely, Nvidia shares fell more than 5%, to $169, bringing the market cap to $4.15 trillion. Since the end of October, when they reached all-time highs, the AI chip maker’s shares have fallen by about 20%, which in terms of market capitalization means a loss of nearly $1 trillion. In fact, Nvidia’s capital value has reached five billion as a result of its growth and alliances with companies in the world of artificial intelligence, such as OpenAI and Anthropic.
Apple is also on the verge of becoming the company with the highest market cap, thanks to its stock rising 42% in the past six months, in the midst of new iPhone models.
In any case, this recovery in Alphabet shares, which had already surpassed Microsoft in market value at the end of last week, strengthens the position of the search giant, which is strengthening its efforts with its new teams to compete with Nvidia’s best-selling artificial intelligence accelerator.
According to Bloomberg Intelligence, third-party providers of comprehensive language models will likely use Google as a secondary supplier of accelerator chips for inference in the near term.
These analysts note that Google Cloud consumption and backlog growth could accelerate compared to other hyperscalers and similar new cloud platforms due to demand from enterprise customers who want to consume TPU and LLM Gemini on Google Cloud.
This site joins the excellent reviews received by Alphabet’s new Gemini AI model, which was launched last week, and which is already considered better than OpenIA technology.
According to specialist firm Vital Knowledge, the Gemini 3’s impressive debut was not only due to the model’s capabilities and performance, but also because Google trained much of this new technology on its own AI chips. These analysts believe that Alphabet could become a competitor to Nvidia.
In his opinion, while Nvidia remains the dominant GPU supplier, its market share will struggle increasingly in 2026, both from custom chips (Google’s TPU and Amazon’s Tranium) and new AMD products.
In this sense, Google launched so-called tensor processing units a decade ago to speed up the Internet search engine and increase its efficiency. These devices were later adapted for machine learning tasks in Google’s AI applications.
Now, Alphabet is closing big deals for TPUs, which could include a potential deal with Meta, suggesting, according to Bloomberg, that they could be a reliable alternative to Nvidia’s AI accelerators for training and running today’s complex, large-scale language models.
About The Author
