Alphabet said on 28 October 2026 that it held $13.3 billion of inventory on 30 September, up from $2.4 billion at the end of 2025. Its second-quarter 10-Q says that inventory “consists primarily of hardware related to TPU systems for sale to enterprise customers and devices, which primarily include the Pixel family of products”.
By our arithmetic, that is 5.5 times the year-end level, and equal to 43% of Google Cloud’s third-quarter revenue of $30.7 billion.
Short explainer · Google TPU brief
A vertical news Short narrated from the published headlines and opening lines of this piece and four related FourWeekMBA pieces on Google.
Business Pill · Free Cash Flow
An evergreen Business Engineer explainer on free cash flow, the cash a company has left after capital expenditures. It explains the general idea only and does not discuss Alphabet.
The key insight: By our arithmetic from Alphabet’s filings, its inventory is 5.5 times its end-2025 level and equal to 43% of Google Cloud’s third-quarter revenue.
Update, 28 Oct: on the earnings call. Google Cloud chief executive Thomas Kurian said “all of the inventory for TPU systems are associated with committed contracts.” Chief financial officer Anat Ashkenazi described the cycle: a signed agreement first shows up in backlog, then “we build the inventory ahead of getting the revenue stream in,” and cash comes in as systems are delivered and billed.
She put future purchase commitments at $924 billion at the end of the third quarter, and said the amounts “reflect the maximum contractual obligations.” The June-quarter 10-Q listed purchase commitments and other contractual obligations of $811.0 billion at 30 June; the two figures may not be defined identically. (Quotes from a caption transcript of the call webcast.)
Alphabet’s third-quarter release says Google Cloud “generates product revenues primarily from the sale of TPU systems”. It also says “TPU systems sales were also a meaningful contributor to revenuegrowth” in a quarter when Cloud revenue rose 103%.
In its first-quarter 10-Q, Alphabet said Google Cloud had “entered into a limited number of agreements to supply multiple gigawatts of TPU hardware to customers who require or provide on-premises infrastructure for specialized, high-scale workloads”. It said then that it expected to begin recognizing revenue from them later in 2026, with the significant majority in 2027.
The same filing adds that, in connection with certain of these agreements, “we have agreed to provide credit backstops to support third-party data centers and power infrastructure”. Its second-quarter 10-Q says Alphabet began recognizing revenue from the agreements in the second quarter of 2026.
Alphabet inventory at period end, $ billions: 31 December 2025, 30 June 2026 and 30 September 2026. Alphabet says its inventory is primarily TPU-system hardware and Pixel devices. The 5.5x multiple and the percentages of Google Cloud revenue are our arithmetic. Sources: Alphabet Q3 2026 release (8-K Ex. 99.1); Q2 2026 Form 10-Q.
The Inventory Line, Quarter by Quarter
Alphabet’s balance sheet shows inventory of $2.4 billion on 31 December 2025, $10.0 billion on 30 June 2026 and $13.3 billion on 30 September 2026. The first-quarter balance sheet did not show inventory as a separate line.
Its cash flow statement shows inventory absorbing $11.0 billion of operating cash in the first nine months of 2026, against $1.2 billion a year earlier. In the third quarter alone the figure was $3.2 billion, against $0.5 billion a year earlier.
Set against Google Cloud revenue, by our arithmetic, inventory equalled 40% of second-quarter Cloud revenue of $24.8 billion and 43% of third-quarter Cloud revenue of $30.7 billion.
The Commitments Behind It
Alphabet’s 10-Qs list purchase commitments and other contractual obligations of $332.4 billion on 31 March 2026, of which $138.0 billion was short-term, and $811.0 billion on 30 June 2026, of which $200.7 billion was short-term. The filings say these primarily relate to costs for technical infrastructure and inventory.
The same filings put Google Cloud’s revenue backlog at $462.3 billion on 31 March and $513.9 billion on 30 June. Alphabet’s third-quarter release and slides do not give a backlog figure.
What a TPU System Is, in Alphabet’s Slides
A slide in Alphabet’s third-quarter presentation describes the product: TPU trays with four proprietary chips on one liquid-cooled board, racks of 16 TPU machines with battery backup, and pods of 144 TPU racks with three optical switch racks. The same slide lists post-sales product support and warranty, and software support for JAX, PyTorch, vLLM and SGLang.
As we read the filings, the inventory line and the TPU agreements move together: Alphabet first separated inventory on its balance sheet in the second quarter, the same quarter it says it began recognizing revenue from the on-premises TPU agreements.
Its cash flow statement shows inventory absorbing $11.0 billion of operating cash in the first nine months of 2026, against $1.2 billion a year earlier.
The first-quarter 10-Q ties the TPU agreements to credit backstops for third-party data centers and power infrastructure, which means part of the exposure sits outside the inventory line.
Sundar Pichai, CEO of Google and Alphabet, Q3 2026 release, 28 October 2026
“We’re also seeing strong demand for our TPU systems across a wide range of customers”
Three Implications
FOR GOOGLE CLOUD Alphabet’s release says Google Cloud generates product revenues primarily from the sale of TPU systems, a hardware line alongside its consumption-based services.
FOR THE BALANCE SHEET Inventory rose from $2.4 billion to $13.3 billion between 31 December 2025 and 30 September 2026, according to Alphabet’s balance sheets.
FOR TIMING Alphabet’s 10-Qs say the significant majority of revenue from the on-premises TPU agreements is to be recognized in 2027.
The Business Engineer Lens
This story maps onto the Business Engineer framework The AI Capex Map.
The framework argues that “the seven-layer AI stack is now governed from the bottom — energy, silicon, compute capacity.”
Alphabet’s filings show one company selling at that bottom layer: TPU systems for customers’ own sites, with inventory of $13.3 billion behind them. The framework does not discuss these figures.
What Is Not Established
The filings do not split inventory between TPU-system hardware and Pixel devices, do not report TPU-system revenue as a separate figure and do not name the customers of the on-premises agreements.
Our 43% compares a balance on one date with a quarter’s revenue, and Google Cloud revenue includes services as well as product sales. It is a scale comparison, not a measure of how fast inventory turns into revenue.
This Week In The AI Supercycle
Execution Fault Lines, Concentrated Demand, and Geopolitical Firewalls
This week’s issue: the AI build is not short of demand for its parts. It is short of cheap money, and its end revenue rests on a narrow base of large, often venture-funded buyers.
Alphabet’s filings describe a new hardware business inside Google Cloud: TPU systems sold for customers’ own sites, revenue recognized from the second quarter of 2026, and an inventory line that grew from $2.4 billion to $13.3 billion in nine months.
Business Engineer Premium
Premium: Alphabet & Microsoft on the Map of AI
Cash conversion, the lengthening backlog, the arms-dealer turn at Layer 3: the full Business Engineer read of both prints.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. Figures come from Alphabet’s third-quarter 2026 results release of 28 October 2026 (SEC Form 8-K, Exhibit 99.1), its third-quarter earnings slides and its Forms 10-Q for the quarters ended 31 March and 30 June 2026. The multiple and percentages are our arithmetic; linked earlier pieces carry their own sources.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
