① Google's release of the Gemini 3 series of large-scale models successfully mitigated the disruptive risks of AI to its search business, enabling it to shift from a defensive to an offensive posture in the AI race. ② HSBC Analyst Paul Rossington believes the Gemini 3 Pro is superior to ChatGPT 5.1 and has raised his target price for Google's parent company, Alphabet, to $370.
HSBC stated that the success of Google's newly released Gemini 3 series of large-scale language models has mitigated potential artificial intelligence risks for the tech giant's search business. (AI) Disruptive risks. At this point, Google has shifted from a defensive to an offensive stance in this AI race.
"We believe the Gemini 3 Pro is a cutting-edge model that outperforms the ChatGPT 5.1, which strengthens our confidence that Google can expand (not just maintain) its share of the search market," HSBC analyst Paul Rossington wrote in a report to clients.
He maintained his "buy" rating on Alphabet, Google's parent company, and raised his price target from $335 to $370, partly due to the performance of the Gemini 3. This implies that the stock could still rise by about 17% from current levels.
Since its public release last month, Gemini 3 has received rave reviews. It surpasses its predecessor in coding, design, and analysis, and outperforms competing models in benchmarks. Media outlets have found it particularly adept at designing websites and basic video games. This allows it to have a wider range of uses beyond encoding.
Even software company Salesforce CEO Marc Benioff has publicly stated that he will abandon ChatGPT in favor of Gemini 3, calling the latter a “crazy” leap in inference, speed, and multimodal capabilities.
Amidst the widespread acclaim for the Gemini 3, other developers are finally feeling the pressure. Reports indicate that OpenAI, the company behind ChatGPT, is developing a new model codenamed "Garlic." On Tuesday, local time, French AI company Mistral also released its latest model series, the Mistral 3.
On the other hand, with the emergence of Gemini 3, its self-developed AI chip The TPU (Tensor Processing Unit) has finally stepped out from behind the scenes and into the spotlight.
Rossington further pointed out that if Google can truly commercialize the TPU, it might create something equivalent to Nvidia by 2030. This represents 10% of revenue and increases Alphabet's market capitalization by approximately 6%.
Last week, reports surfaced that Google is marketing its TPUs to customers, and Meta is considering spending billions of dollars to purchase Google TPUs for use in its data centers. Construction has begun. This marks a significant shift in Google's strategy, as its TPUs were previously primarily used within its own cloud data centers .
Previously Morgan Stanley Analysts say that given Nvidia’s projected shipment of approximately 8 million GPUs (graphics processors) in 2027, Google’s forecast of selling 500,000 to 1 million TPUs annually is “not unreasonable” should production capacity allow.
Rossington concluded, "The complete AI strategy is making substantial progress through initiatives such as the release of Gemini 3 and the potential external commercialization of the TPU. The Gemini-driven AI initiative and its commercialization results will inject vitality into Google's long-term prospects. "
(Article source: CLS)
