Google Falls Short In Cloud, Overshoots On Capex

Alphabet on Tuesday turned in a set of results which, at first glance anyway, looked underwhelming. As ever, it's important to note that "underwhelming" in the context of the mega-caps is a highly relative term. Specifically, it's relative to consensus, and pretty much only to consensus because outside the realm of Wall Street expectations, there's nothing "underwhelming" about, say, $96.5 billion. That's how much revenue Alphabet raked in last quarter. Sales growth was 12%, the slowest since

Join institutional investors, analysts and strategists from the world's largest banks: Subscribe today

View subscription options

Already have an account? log in

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

4 thoughts on “Google Falls Short In Cloud, Overshoots On Capex

  1. Revenue missed and topline growth was the slowest since 3Q23, if I have it right. Services rev +10.2%, a deceleration from 3Q +12.5%. Cloud rev +30.1%, a big deceleration from 3Q +35%. Claimed capacity constraint in AI and soaring capex is consistent, but would AI constraint explain such deceleration in total Cloud or is the non-AI revenue decelerating? Unclear. The deceleration was mostly in the US. Operating margin improved very nicely, but was flattered by charges last year. 1Q25 has FX, Leap Year, 4Q election ad headwinds. Cloud growth during 2025 to be “variable” with lumpy capacity adds. Expense headwinds from accelerating depreciation and some headcount growth. Obligatory assurance on managing the ($75BN of) capex “responsibly”. Said Gemini and TPU lead in efficiency, sidestepped an invitation to brag more about TCO. The call seemed kind of low-energy compared to META’s “2025 is going to be a big year!” vibe. I think estimates should go down a little for 1Q if not 2025, not much but anyway not going up. Still, seems positive for GOOG’s AI chip supplier, can debate if better for AVGO or for NVDA, and the capex-unchained theme.

    1. Watching after hours price action in GOOG and AMD tonight, it seems like at least some investors are finally starting to ask “where’s the beef?” when it comes to endless monster capex on AI infrastructure.

      The 18 month chart on AMD is especially noteworthy. The rise and fall of the stock based on AI hype mirrors what we’ve seen in metaverse, blockchain and weed stocks in recent years. Even though the company managed to bring AI chips to market which offer strong performance relative to Nvidia’s offerings at a much lower price.

      Where’s that dead canary illustration our Dear Leader has used from time to time?

      1. AMD’s AI GPU revenue missed buyside revenue expectations for 2024. AI GPU rev grew less sequentially than server CPU rev did in 4Q, are guided to be flattish for 1H25 as compared to 2H24, with renewed sequential growth not until 2H25, and company isn’t disclosing targets for 2025 beyond “strong double digits”. Basically the fear is if momentum for MI300/325 has stalled and won’t re-accelerate until MI350 ramps in, supposedly, mid-year 2025 . . . or maybe until MI400 comes in 2026 . . . or . . . when?

        Too much uncertainty. Meanwhile consensus has NVDA’s data center revenue growing around +10% sequentially every quarter from now through at least end of 2025.

        Shame because AMD is doing very well – given the environment, better than very well – in server CPU and client CPU. Revenue is growing, margins climbing, share growing. But all that matters to the market is AI GPU.

NEWSROOM crewneck & prints