Alphabet has been re-rated as a beneficiary of artificial intelligence infrastructure, sending its shares up about 160 percent over the past year. In after-hours trading on May 10, it briefly surpassed Nvidia by market capitalisation.

CNBC reported that the market is focusing on Google’s broad lineup, from AI models and cloud to in-house semiconductors, as well as distribution channels spanning search, YouTube and Android.

Early in the AI boom, Google was seen as lagging rivals. More recently, its business structure spanning chips, models, infrastructure and distribution has emerged as a strength. Gene Munster (진 먼스터) of Deepwater Asset Management named Google as one of the AI companies in the most advantageous position and viewed its profitability positively.

Wall Street is also turning more positive after the earnings release. JPMorgan named Alphabet its top pick within the technology sector, citing accelerating growth at Google Cloud and an order backlog that has grown to $462 billion. Mizuho raised its price target, judging that market expectations over the next 2 years still underestimate Google Cloud revenue and operating profit.

Market attention is focused in particular on the cloud business and in-house AI chips. Alphabet posted a market value of $4.8 trillion based on its weekly closing price, closely chasing Nvidia at $5.2 trillion.

Behind the narrowing gap in market value is news of a large contract between Anthropic and Google Cloud. Reports said Anthropic plans to invest $200 billion over the next 5 years for 5 gigawatts of computing resources on Google Cloud.

There are also concerns about the order structure. Gil Luria (길 루리아) of D.A. Davidson pointed out that most of the increase in the order backlog may have come from the single Anthropic deal. If the $200 billion commitment was reflected in Alphabet’s disclosed cloud order backlog, it could account for more than 40 percent of future contract revenue, by that calculation.

Such concerns are cited as a common issue for other large cloud operators. Luria said the combined cloud order backlogs of Microsoft, Oracle, Amazon and Google are close to $2 trillion, and that a significant portion is tied to commitments from OpenAI and Anthropic. He said the structure is that the 2 AI companies raise funds from the same large technology companies and then buy cloud and chip services from them.

There are also counterarguments. Munster said the deal instead shows that the AI market is still at an early stage. He said use cases are currently limited but demand for computing is rising exponentially and that Google will ride that trend. He added that even if a specific customer falters, many AI companies will emerge over time to replace it.

Google’s differentiator is cited as custom semiconductors. Mizuho estimated that by 2027, about $61 billion of Google Cloud’s order backlog could come from sales of TPUs, or tensor processing units. Most of that revenue is highly likely to be recognised next year. That is why Google is emerging as another choice for investors seeking alternatives to Nvidia for AI hardware investment.

On the other hand, some have raised the view that a significant share of demand for in-house chips may come from within affiliated investment ecosystems. Luria said Google and Amazon emphasise demand for their chips, but many parts of it may be captive demand coming from portfolio companies and may not be organic demand.

A key point to watch will be whether large-scale investment can be recouped. Alphabet has set this year’s capital expenditure at up to $190 billion, more than double the 2025 level. The market is looking to see whether the expanded investment will translate into sustainable AI profits. At Google I/O, expected within 2 weeks, the Gemini-based agent strategy and monetisation plans across the broader AI ecosystem are expected to be key checkpoints.

Ultimately, Alphabet has quickly shifted from an image as a latecomer in AI to a beneficiary of infrastructure. But with the share price already reflecting expectations for the future to a significant extent, Google has entered a phase in which it must prove performance in its cloud order structure, customer concentration risks and large-scale capital expenditure.