Azure's contribution to cloud computing revenue has exceeded 50% for the first time, and OpenAI's enterprise users have exceeded 11,000.
After the market closed on July 25th, local time, Microsoft released its fourth-quarter report for 2023.
According to the report, Microsoft's fourth-quarter revenue and profit both exceeded expectations, but the growth rate of its "cash cow" business Azure slowed down. In addition, the investment in AI has caused Microsoft's capital expenditure to exceed expectations, leading to a nearly 4% drop in Microsoft's stock price after hours.
During the fourth-quarter earnings conference call held later, Microsoft's Chairman and CEO, Satya Nadella, stated that Microsoft remains focused on three key priorities: helping customers maximize the value of their investments through the breadth and depth of Microsoft's cloud, continuing to lead the AI platform by injecting AI at every layer of the technology stack, and improving operational leverage.
He emphasized that the sales of Azure cloud computing services account for more than half of Microsoft's $110.9 billion cloud computing revenue in the 2023 fiscal year, marking the first time Azure has accounted for the majority of Microsoft's annual cloud computing business.
Microsoft's CFO, Amy Hood, stated that although there is strong demand for Azure AI services at present, the contribution of AI services to Azure's revenue is currently only about 1 percentage point. As Microsoft accelerates its investment in cloud infrastructure, the impact of AI on Microsoft's revenue will be concentrated in the second half of the 2024 fiscal year.
Explanation of Capital Expenditure: Strong Demand Signals
Nadella and Hood addressed questions regarding the increase in capital expenditure.
Hood stated that Microsoft's acceleration in AI infrastructure is extensive, both in terms of data centers and physical infrastructure construction, as well as investments in CPUs, GPUs, and network devices. This represents an overall increase in capabilities, driven by both commercial cloud demand and the need for AI:
I think if you look back at the 2023 fiscal year, you'll see two accelerations, the normal Azure workloads plus some AI workloads. So, that's why I often comment that it's both the overall commercial cloud demand and the AI capability build-out.
I think, in the long run, I think it's always good to think about it, we build the commercial cloud business growing at 22% annually. And then capital expenditure is growing at about 23%, 24% as well. So, in a way, it's a substitution of capital plus some new capital to drive new growth.
Hood emphasized that the higher-than-expected capital expenditure is due to Microsoft's recognition of "demand signals":
We have seen very strong demand signals, and they remain strong.
Our products do have scalability in potential markets, reaching new budget pools. So I discussed these investments with CIO - it's a growth opportunity.
Revenue is a result, of course, and it is necessary. That's why, in a sense, we spend more time talking about investments because we see the demand signals.
Nadella added that the boost from AI to Microsoft will gradually become evident in the second half of the new fiscal year:
Assuming global technology spending increases from 5% of GDP to 10% - which may accelerate due to the AI wave - the question is how much of it will flow into different parts of our commercial cloud and how competitive we are in those parts.
We already have productivity tools, then we have all the communication and collaboration tools, and now we have Microsoft 365 and CoPilot as the third pillar.
In the second half of the new fiscal year, we will start to see some real revenue signals from it. As for cloud migration, we are still in the early stages, so there is still a lot of work to be done. But we do see very strong demand signals, and they remain strong.
AI gross margin growth faster than cloud
When asked how to maximize the value of generative AI, Nadella stated that Microsoft will focus on helping customers build generative AI applications on Azure AI:
We want to be able to help customers quickly build generative AI applications on top of Azure AI, and we have our own resources to help them do that.
In terms of gross margin, Hood expects it to grow faster than during the period of cloud development:
I expect gross margin to evolve over time, just like the previous cloud transformation. We expect the workloads to be different. We start from a different place, with more shared platforms, which allows us to expand gross margin faster than last time. We do expect gross margin to expand faster than during the previous cloud expansion. So you will see capital expenditure accelerate in the fourth quarter and then accelerate again in the first quarter, and we have discussed the situation for the rest of the year.
I think the real focus here is being able to proactively meet the demand curve, focusing on transformation, growth, and gross margin, and providing operating leverage.
Azure contributes over half of cloud revenue
Nadella stated during the conference call that the sales of Azure cloud computing services account for over half of Microsoft's fiscal year 2023 revenue, marking the first time Azure's share of cloud revenue has exceeded 50%.
Among them, Azure AI service revenue has grown by about 50%, contributing approximately two percentage points to Azure's revenue growth of 27%. The demand for Azure OpenAI services continues to grow rapidly. Azure OpenAI currently has 11,000 corporate clients, a significant increase compared to the 4,509 reported in mid-May. On average, 100 new clients are added each day this quarter.
Speaking of artificial intelligence, Nadella highlighted several new AI products and features, including GitHub Copilot, Power BI Copilot, Windows version Copilot, and Microsoft 365 Copilot.
Nadella stated that feedback indicates these suites significantly improve work efficiency. For example, GitHub Copilot can increase developer productivity by 40% to 50% or even more, generating higher-quality code. Microsoft hopes that Microsoft 365 Copilot will also validate this conclusion.
The easiest way to harness the value of generative AI is to adopt solutions like GitHub Copilot. In a sense, it is a straightforward choice for all software developers in any organization - a productivity boost.
It applies to banks, retailers, and software companies alike. So, we have seen very positive adoption and efficiency data.
He mentioned that Microsoft Fabric has attracted over 8,000 registered users in just one month. GitHub Copilot has over 27,000 corporate clients, doubling its growth rate. The AI-driven version of Power Platform has been adopted by over 63,000 corporate clients, experiencing a 75% increase. Bing users have engaged in over 1 billion conversations and created over 750 million images using Bing Image Creator. Microsoft Edge has consistently gained market share for the ninth consecutive quarter.