Despite its Q3 backlog jumping 49% to $364 billion, AWS finished a distant fourth in the latest Hyperscaler Backlog Sweepstakes as the world’s top AI factories — Google Cloud, Oracle, Microsoft, and AWS — have generated a combined backlog of $2.1 trillion, up a whopping 23% in just four months.
While there’s no question that massive and unprecedented orders from AI vendors form a big chunk of that $2.1 trillion in fully contracted business not yet recognized as revenue, the hyperscalers are all reporting that the AI Revolution has triggered robust growth from businesses in every industry and every region of the world.
Before we get into some of the details, let’s consider for a moment some of the implications of this $2.1 trillion backlog:
- Astonishing customer demand and hyperscaler growth at levels the business world has never seen.
- A total and complete repudiation of the absurd “AI bubble” paranoia.
- A clear indication that the challenge is not one of frantically trying to create demand but rather the need to fulfill it with some of the largest and most technologically advanced facilities and equipment the world has ever seen.
- The remarkable evolution from what were primarily “software” companies into multifaceted hubs of hardware, software, networking, cybersecurity, energy generation, power and construction innovation, water conservation, and fresh thinking on utility payments.
- Accordingly, the close convergence of the formerly disparate industries of technology, energy, power generation, public utilities, and nuclear fusion.
- New business models for the hyperscalers who must juggle and solve wildly complex equations involving skyrocketing demand, supply-chain precision, new revenue models, and the ultimate synchronization of demand and supply with levels of availability unlike any in human history.
- And, lest we forget, the four hyperscalers all admit that in spite of their plans to spend somewhere in the range of $650 billion this year in their herculean efforts to meet the absolutely wild levels of demand that exist, they will all continue to be capacity-constrained for some time.
By the Numbers: Handicapping the Hyperscalers
Using the latest quarterly financial results from Google Cloud, Oracle, Microsoft, and AWS, I’ve assembled below the current backlog totals and growth rates for the four hyperscalers. While Oracle’s numbers were released June 10 as part of its Q4 disclosures, I’ll update these in late July or early August when calendar-Q2 numbers are released for Google Cloud, Microsoft, and AWS.
| Backlog | RPO Total Backlog | RPO Growth Rate |
|---|---|---|
| Oracle | $638 Billion | 363% |
| Microsoft | $627 Billion | 99% |
| Google Cloud | $462 Billion | 93% |
| AWS | $364 Billion | 49% |
| Total | $2.091 Trillion |
Final Thought
To offer a bit more perspective on just how extraordinary these numbers are, let me close with the same detail I used to open this piece: in a brilliant Q1 performance, AWS boosted its backlog by 49% to $364 billion.
But, while no companies outside of the tech industry have ever posted numbers of that scale, within the world-shaping hyperscaler business, that performance by AWS was only good enough for fourth place — and a distant fourth at that.
My friends, we are living in remarkable times — we would all do well to be aware of that good fortune, and enjoy it.
