The Alpha In Infrastructure

Similar to what we shared last quarter, we continue to see the diminishing impact of cost optimizations. And as these optimizations slow down, we're seeing more companies turning their attention to newer initiatives and re-accelerating existing migrations.

Customers are also excited about our approach to generative AI. It's still relatively early days, but the revenues are accelerating rapidly across all three layers, and our approach to democratizing AI is resonating well with our customers. We have seen significant interest from our customers wanting to run generative AI applications and build large language models and foundation models, all with the privacy, reliability, and security they've grown accustomed to with AWS.

If you look at the GenAI revenue we have, in absolute numbers, it's a pretty big number, but in the scheme of a $100 billion annual revenue run rate business, it's still relatively small, much smaller than what it will be in the future where we believe we're going to drive tens of billions of dollars of revenue over the next several years. But it's encouraging how fast it's growing and our offering is resonating with customers.

with Azure AI, we provide access to the best selection of foundation and opensource models, including both LLM and SLMs all integrated deeply with infrastructure, data, and tools on Azure. We now have 53,000 Azure AI customers, over one-third are new to Azure over the past 12 months.

Azure and other cloud services revenue grew 30% and 28% in constant-currency, including 6 points of growth from AI services. Both AI and non-AI Azure services drove our outperformance.

For developers building Gen AI applications, we offer Vertex AI, a comprehensive enterprise AI platform. It helps customers like Deutsche Telekom and Moody's discover, customize, augment, and deploy over 130 Gen AI models, including Palm, Med-Palm, Sec-Palm, and Gemini, as well as popular open-source and partner models.

Vertex AI has seen strong adoption with API requests increasing nearly 6x from H1 to H2 last year. Using Vertex AI, Samsung recently announced its Galaxy S24 series smartphone with Gemini and Imagen 2, our advanced text-to-image model. Shutterstock has added Imagen 2 to their AI image generator, enabling users to turn simple text prompts into unique visuals. And Victoria's Secret & Co. will look to personalize and improve the customer experience with Gemini, Vertex AI Search, and Conversations.

Meta has a long history of building new technologies into our services, and we've a clear long-term playbook for becoming leaders. And there are a few key aspects of this that I want to take some time to go through today. The first is world-class compute infrastructure. And I recently shared that by the end of this year, we'll have about 350,000 H100s, and including other GPUs, that'll be around 600,000 H100 equivalents of compute.

Going forward, we think that training and operating future models will be even more compute intensive. We don't have a clear expectation for exactly how much this will be yet, but the trend has been that state-of-the-art large language models have been trained on roughly 10x the amount of compute each year