No AI hyperscaler hyperscare here

A quartet of Magnificent 7 (Mag 7) companies reported on their quarterly earnings results recently (29 April). Microsoft, Amazon and Alphabet (Google Cloud) are the three largest hyperscalers, representing broad demand for AI use cases, while Meta is a leading use case of AI monetisation in its own business. Hyperscaler revenue is likely to be the most important lead indicator of AI returns and monetisation this year – recent earnings results demonstrated accelerating growth, strong backlogs and product expansions.

Varied reactions to the earnings reflect our long-held belief that the Mag 7 are not a monolith (vastly different businesses, margins, valuations and growth rates). But we also note that in the results from these four companies, there were more commonalities than differences:

1. Revenue is accelerating across all the companies with capacity rising to meet surging demand, as is backlog (the booking figure reported on cloud divisions showing future sales). The backlog of Amazon, Alphabet and Microsoft announced exceeded US$1.4 trillion.
2. Agentic AI is inflecting as the shift from training workloads to inferencing workloads drives an inflection/ exponential step up in demand for compute power, storage and networking.
3. AI hyperscalers continue to be supply constrained in fulfilling demand and are responding to their backlog with higher capital expenditure (capex).
4. Capex is shifting higher still – the four companies have committed cumulatively to more than US$700 billion in capex in 2026.1 A combination of rising costs (eg. memory) and the strength of demand have led management teams to indicate ( to various degrees) that this capex cycle will be stronger for longer beyond 2027). This is key for the AI infrastructure (e.g. semiconductor supply chain etc.), which looks to have already been partially priced into valuations. We expect 2027 capex to grow further.

What recent earnings reveal about the AI build out

The latest earnings results were positive for the build out of the full AI technology stack and ecosystem. Below we highlight some of the nuances and key points from each of the companies:

Amazon (NASDAQ: AMZN) 

We’ve never seen a technology grow as rapidly as AI, Amazon is already a leader, and companies continue to choose AWS for AI.
- Amazon CEO Andy Jassy

Amazon delivered impressive results. Given Amazon Web Services (AWS) is the largest cloud provider, investors had expected strong growth from the division – there was no disappointment with AWS revenue growth accelerating to 28% on a US$150 billion revenue run rate.

The CEO articulated strong visibility on attractive return on invested capital (ROIC) from rising capex. While Amazon did not provide explicit 2027 capex guidance, demand strength was evident in its US$364 billion backlog (of contracted future revenue), which excludes the recently announced north of US$100 billion Anthropic commitment towards AWS technologies over the next ten years.2 Amazon’s differentiated approach to customer semiconductor chips has continued to gain traction, with its Graviton CPU now being sold to third parties, including Meta, and its Trainium AI chip now at a US$20 billion run rate. The retail groceries division demonstrated improving margins and strength with 15% year-over-year unit growth –outpacing fulfilment‑network cost growth. Broad‑based momentum across the US, Europe, and Brazil may help reassure investors concerned about macroeconomic pressures.

Alphabet (Google) NASDAQ: GOOGL 

Our enterprise AI solutions have become our primary growth driver for cloud for the first time in Q1.
- Alphabet CEO Sundar Pichai

Alphabet has the least mature, fastest-growing cloud business, so investors have looked for evidence of its scale potential. The company announced very strong growth in Google Cloud revenues of 63%, with associated operating margins having almost doubled. There was also disclosure of an exceptional US$462 billion backlog (up 90% year‑on‑year), with Google expecting to work through 50% of that over the next two years. This is one of the largest backlogs ever reported, reflecting both a doubling of new customer acquisitions and deepening relationships with existing customers.

Alphabet is unique in its fully integrated, full‑stack AI approach — spanning frontier models, Google Cloud applications, cloud infrastructure, and its own tensor processing unit (TPU) silicon. Unusually, management commented on 2027 capex, signaling a significant increase in spending and confidence in continued AI monetisation. AI also benefited the core Google Services business, which grew 16% year-on-year in its first quarter. Meanwhile, Waymo, its autonomous driving division, is now serving 500,000 rides per week, illustrating another future growth vector in physical AI.

Microsoft (NASDAQ: MSFT)

Weekly engagement is now at the same level as Outlook, as more and more users make Copilot a habit.
- Microsoft CEO Satya Nadella

Investors were seeking a clear narrative on how the company’s capex would be balanced between internal operations and its third-party cloud/AI business.

Microsoft again reported strong results, with the Azure cloud business growing 40% in the quarter from a year ago, although management reiterated that growth continues to be constrained by compute availability. The AI business surpassed US$37 billion in annual recurring revenue (ARR), up 123% year-on-year, attributed to customers building and running AI solutions on the Azure platform. However, growth of the AI business was lower versus Alphabet and Amazon, while concerns about high exposure to Open AI and associated huge capex and monetisation remain.

Meta (NASDAQ: META) 

People will be more important in the future, not less.
- Meta CEO Mark Zuckerberg

Meta reported strong advertising growth, with revenues for the quarter up 31% year‑on‑year. The weaker market response to Meta’s earnings despite its 30% revenue growth (the strongest of all the four) reflects the fact that Meta has no backlog to report and hence there is less visibility on the future returns of its capex. A small decline in Family of Apps (Facebook, Instagram, WhatsApp, Messenger, Threads) daily active users was attributed to internet shutdowns in Iran and WhatsApp being blocked in Russia. Engagement trends remained strong, with improved monetisation for advertisers. Video engagement reached all‑time highs across Facebook and Instagram, driven by ranking improvements. Meta’s innovation in advertising saw higher adoption of tools such as the Value Optimization Suite and Partnership Ads. During the quarter, Meta launched Muse – its newest family of large language models LLMs, which the CEO described as being integrated into personal AI tools via Meta AI, as well as applications in shopping and social content. However, these initiatives drove a capex increase of US$10 billion. Q2 guidance was modestly raised, while full‑year operating expense guidance was maintained. There was no guidance on margins improvement from the recently announced 10% workforce cut.

Conclusion

Overall, the latest earnings results from the four of the Mag 7 showed accelerating momentum in AI demand and increasing monetisation potential from the large cloud platforms. However, this continues to be constrained by capacity as evidenced by their backlog numbers. This is clearly not a zero‑sum game: all major hyperscalers are seeing strengthening demand as agentic AI adoption creates another inflection point for compute and storage requirements.

We believe the build out of infrastructure and applications for AI is a multi-year process, as we saw during the internet and mobile compute tech waves where significant investment is required to realise its full potential.