GPU Depreciation and the Hidden Cost Behind AI Financial Statements
GPU purchases are capital investments, but their cost reaches the income statement through depreciation. AI infrastructure makes asset life assumptions financially important.
Buying a GPU is not the same as expensing a GPU
The first step is to define the metric precisely because finance terms that sound intuitive often have specific accounting boundaries. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. CoreWeave reported $2.454 billion of depreciation and amortization in 2025 and another $1.393 billion in Q2 2026 alone. [1]
A useful analytical habit is to separate operating metrics from accounting statements. Operating metrics can be excellent leading indicators, but they often omit financing structure, depreciation, stock compensation, tax, working capital, or the capital needed to sustain growth.
Capital expenditure becomes expense gradually
Labels are useful only after the underlying calculation is understood.
Depreciation spreads today’s capital bill across future revenue
The current AI market provides unusually vivid evidence because companies are scaling revenue, compute, and capital commitments at the same time. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. CoreWeave places depreciation related to power systems and computing infrastructure inside cost of revenue and technology-and-infrastructure expense. [2]
The second habit is to ask what happens when usage doubles. If revenue doubles while direct serving cost rises almost as fast, scale may improve the headline without creating much operating leverage. If cost grows much more slowly, the same growth can produce powerful margin expansion.
Idle hardware still depreciates
A dramatic growth rate can coexist with weak unit economics.
CoreWeave makes the accounting impact unusually visible
The mechanism matters: the same headline number can imply very different economics depending on what sits above or below it in the financial statements. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. Epoch notes that the capex required for a frontier cluster can be many times the cost of a single model-training run because the hardware is expected to serve workloads for years. [3]
A third distinction is timing. Accounting can spread some costs across years, recognize some revenue over contract periods, and exclude certain items from management-defined measures. Cash, however, moves when suppliers, employees, lenders, and infrastructure vendors are actually paid.
Obsolescence can outrun the original utilization plan
Cash and accrual accounting answer different timing questions.
Hardware generations can change faster than accounting lives
AI intensifies the issue because model serving, infrastructure, research, and strategic financing introduce costs that ordinary software companies could often ignore. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. Nvidia’s rapid transition from Hopper to Blackwell and Blackwell Ultra illustrates how quickly AI hardware generations change. [4]
AI also makes capital structure part of product strategy. Companies with wealthy parents, strategic cloud partners, customer prepayments, or public-market access can finance expensive capacity years before a smaller competitor could. That can alter both market share and reported economics.
High utilization is an accounting and operating advantage
AI scale magnifies small accounting assumptions into large valuation differences.
Utilization determines whether depreciation becomes productive cost
Comparisons are useful only when the underlying definitions match. Two companies can use the same label while measuring different economic realities. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. Oracle announced plans to raise $45 billion to $50 billion in 2026 to finance infrastructure needed for contracted cloud demand, showing the capital scale behind AI capacity. [5]
Definitions become especially important in private markets because investors often receive operating metrics without a full public filing. ARR, adjusted operating income, or gross margin may be informative, but an outsider may not see every exclusion or balance-sheet obligation.
Training clusters must earn money after the flagship run ends
For investors, the important question is how the metric connects to future cash generation rather than whether the headline number looks large. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. CoreWeave reported $2.454 billion of depreciation and amortization in 2025 and another $1.393 billion in Q2 2026 alone. [1]
The best comparison therefore follows the money from customer payment to gross profit, operating expense, interest, tax, capital expenditure, and finally free cash flow. A metric is useful to the extent that it helps explain one part of that chain without pretending to be the whole chain.
Cloud providers finance capacity years before all revenue is recognized
Technology history repeatedly shows that growth metrics become less persuasive once markets mature and financing is no longer abundant. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. CoreWeave places depreciation related to power systems and computing infrastructure inside cost of revenue and technology-and-infrastructure expense. [2]
Valuation adds a future-tense layer. Markets can rationally pay for growth before current profit exists, but the price ultimately assumes that future revenue will convert into margins and cash after all required investment. The more capital intensive the model, the harder that conversion becomes.
AI profitability depends on turning depreciating silicon into recurring cash
The durable interpretation is therefore the one that survives reconciliation to revenue, expense, cash flow, and capital requirements. The economics of AI infrastructure depend heavily on how billions of dollars of servers and GPUs are depreciated across time while hardware improves rapidly. Epoch notes that the capex required for a frontier cluster can be many times the cost of a single model-training run because the hardware is expected to serve workloads for years. [3]
For CodeHistory, the larger historical point is that AI has not abolished finance. It has made old concepts—revenue quality, depreciation, operating leverage, dilution, cash flow, and cost of capital—more important because the sums involved are so much larger.
Works Cited
- 01CoreWeave FY2025 10-K sec.gov
- 02
- 03
- 04Nvidia Q2 FY2027 10-Q sec.gov
- 05
CodeHistory is a living archive. Citations document the evidence used for this edition; later evidence may refine the account.
Submit a research lead