I believe it is time again to write about the state of Gen AI. We are at the middle of the final calendar quarter of the year and just had another episode of quarterly earnings reports and analysts reports. I continue to find it fascinating how the Gen AI developments are shining in so many different colors and variations of brightness – but still far away from loosing its shininess. So let’s dive right in – and bare with me. I am attempting to look at the health of Gen AI from the ground up.
The Gen AI engine
The first one to mention of course in any Gen AI state of the nation review is Nvidia – and I was blown away by their latest report. Revenue was $30.04 billion, up 15% from the previous quarter and up 122% year-over-year. GAAP earnings per diluted share was $0.67, up 12% from the previous quarter and up 168% year-over-year. – 122% yoy revenue growth and 168% earnings growth… when was the last time you have seen such numbers from an established company (and in the end Nvidia is already more than 30 years old)? And with these dimensions. The margin is not in any way shape or form shy of the success-story: Gross margin was 75.1% on a GAAP basis and 75.7% on a non-GAAP basis. And this has to be seen in a market where the Hopper architecture is at the end of its life-cycle and the upcoming Blackwell architecture required significantly more investments and time than originally planned.
Another Gen AI “engine-company” is of course TSMC – how did they perform? – TSMC reported revenue of $23.5 billion for Q3 2024, which represents a 36% increase year-over-year, and a net profit that surged by 54% compared to the same quarter last year, reaching $10.05 billion. And last but not least, the gross margin is at 57.8%, increased by 4.6 percentage points.
So the engine makers of our Gen AI assessment are doing very well and do not see any reason to slow down nor are they slowed down by anyone or anything. This is impressive.
The Gen AI engine room
Let’s look a bit wider into the engine room and talk about the companies that put all of these impressive chips into their engine rooms, aka data centers.
Let’s start with Microsoft. Microsoft reported revenue increases of 16% year-over-year to $65.6 billion, an operating income growth of 14% to $30.6 billion. Microsoft showed continued excellent cloud performance with revenue increases of 20% to $24.1 billion, with Azure and other cloud services contributing a 33% revenue growth. But this is not yet the interesting fact – Microsoft also separately reported AI performance and AI services contributed 12 percentage points to Azure’s 33% revenue growth.
Let’s turn to Google, so in essence Alphabet. Alphabet’s revenues increased by 15% year-over-year to $88.3 billion, net income rose by 34% to $26.3 billion, with earnings per share (EPS) increasing by 37% to $2.12. This is as impressive and of course both revenue and earnings surpassed analyst estimates. At Alphabet it is a bit more difficult to understand AI contributions as Google utilizes AI across a wide array of their services. But drilling into Google Cloud as this is the B2B relevant number, it showed impressive growth with revenues increasing by 35% to $11.4 billion. Looking into this, a question came up in my mind that needs to be discussed at a separate, later stage: Where and how will the most money be generated from Gen AI? And I think we need to look very closely into the marketing sector from both sides, creating marketing materials automatically or more independently via Gen AI but also Gen AI will play a significant role in increased targeting. And as this part of our economy already pays a significant contribution to today’s technologies, it is save to assume this will build a significant contribution to the overall Gen AI revenue.
Last but for sure not least let’s look at AWS. Already the biggest cloud provider, the one that brought cloud computing on the enterprise table. Amazon’s third quarter 2024 earnings report, released on October 31, 2024, showcased impressive financial performance and growth across various segments. AWS sales increased 19% year-over-year to $27.5 billion, with operating income of $10.4 billion. And Andy Jassy added that AI revenue as part of AWS revenue was growing at triple-digit-rates.
It can be fairly said that the Gen AI engine room is still running on all cylinders and no slow down is in sight.
The Gen AI fuel
And of course we shouldn’t forget the fuel that the LLMs bring to the engine and engine rooms: OpenAI and Anthrophic but also others like Llama or Mistral.
OpenAI has successfully closed a massive $6.6 billion funding round, marking one of the largest venture capital investments in history. This funding round has propelled OpenAI’s valuation to an impressive $157 billion, nearly doubling its previous valuation of $86 billion from earlier this year.
Anthropic, a prominent AI startup and competitor to OpenAI, is currently in early discussions with potential investors about a new funding round that could significantly increase its valuation. According to recent reports, Anthropic is targeting a valuation between $30 billion and $40 billion. This would represent a substantial increase from its previous valuation, essentially doubling the company’s worth. However, it’s important to note that these talks are still in the preliminary stages, and it’s uncertain whether investors will agree to such a high valuation.
In addition, there are still many new AI Start-ups in the making – many of these from former Open AI celebrities like Ilya Sutskever, who after leaving OpenAI, co-founded a startup called Safe Superintelligence. Also Mila Murati, OpenAIs former CTO is in talks to raise capital for a new AI start-up. And not to forget that Antrophic is also one of those.
Looking at all three infrastructure elements of Gen AI, it is fair to say that the boom is nowhere close to being over. And quite frankly with Mr Trump being the new president and his support for Elon Musk I could imagine that additional money and freedom will be added into this system.
But for how much longer will the customers of this infrastructure continue to throw money at it before they are required to show to their shareholders a return on invest? And who could actually do that in a decent timeframe? But this is for another post.
Leave a Reply