Wire flash
Goldman Sachs: Six Tech Giants Need to Earn $1.42 Trillion More by 2028-2030 to Justify 15% ROIC on AI Compute Bets
Editorial responsibility
- No named human review is recorded for this page.
- Source reporting is collected, normalized, translated or condensed automatically when needed.
- Automatically published source-backed update
A Goldman Sachs research report analyzes the economic scale required for major US hyperscale cloud firms to justify their massive AI capital expenditures. The report focuses on the core market debate: whether the huge investments in AI computing power by companies like Alphabet, Amazon, Microsoft, Meta, Oracle, and SpaceX will yield reasonable returns. Goldman Sachs built a quantitative framework to stress-test the returns on computing investments for 2026-2027. The analysis estimates that these six firms need to generate a combined cumulative revenue of approximately $1.42 trillion between 2028 and 2030 to achieve a 15% annualized return on invested capital (ROIC) on their second-phase AI investments. The report notes that while short-term returns will be pressured by the upfront spending, the long-term outlook is positive, driven by increasing AI adoption and a shift from experimentation to implementation. It also highlights the emergence of consumer AI agents as a key future revenue driver.
Source report
Goldman Sachs has published a research report on the U.S. technology sector, aiming to quantify the scale of AI economics required to justify the return on invested capital (ROIC) for hyperscale cloud providers' AI capital expenditures. The report addresses a central market debate: whether the massive capital outlays by leading U.S. hyperscalers on AI computing infrastructure can generate reasonable returns in the future.
The firm has constructed a quantitative framework to stress-test revenue generation from 2026–2027 compute investments across six companies—Alphabet, Amazon, Microsoft, Meta, Oracle, and SpaceX—providing a benchmark for assessing the profitability outlook of AI infrastructure spending.
Key Drivers of the Capex Surge
Goldman Sachs notes that surging token consumption from large language models has driven a dramatic shift in capital intensity among major U.S. hyperscalers, as they adapt to structural changes in compute demand. The current wave of heavy investment is driven by two factors:
- Real-time demand signals, as the market is currently in a state of supply-demand imbalance.
- The need to meet multi-year compute requirements from core customers.
Emerging Market Consensus
Following the recently concluded Q2 earnings season, investor understanding of this theme has deepened significantly. Based on earnings data and management commentary, two points of consensus have emerged:
- Hyperscalers can generate substantial—and even above-expectation—returns from their 2023–2025 capex cycles, through incremental revenue and operating cash flow.
- Driven by tight compute supply and accelerating end-user demand, cloud providers are entering a new investment cycle (2026–2027) , with current compute service pricing significantly above Goldman Sachs' assumed long-term benchmark.
As a result, the market debate has shifted to a longer-term question: How much return will the 2026–2027 capex cycle generate in the 2028–2030 period? And what benchmark does the ROIC from the earlier cycle provide?
The Analytical Framework
Goldman Sachs has developed a framework to estimate the scale of AI market growth required for hyperscalers to achieve a baseline ROIC on current capex levels. The core analysis asks: What revenue threshold must be reached for the second-phase (2026–2027) AI compute investments by leading U.S. hyperscalers to achieve a 15% annualized ROIC?
Key Assumptions
- Average upfront investment per gigawatt (GW) of compute capacity: ~$420 billion
- Allocation: 70% to compute hardware, 30% to facility infrastructure
- Conservative depreciation rules, plus ongoing operational cost assumptions
Key Findings
The six major U.S. hyperscalers (Alphabet, Amazon, Microsoft, Meta, Oracle, SpaceX) would need to generate a cumulative ~$1.42 trillion in revenue from 2028 to 2030—equivalent to approximately $11.6 billion per GW per year—to cross the 15% ROIC threshold.
Short-Term Pressure, Long-Term Promise
While the sharp expansion in short-term capex will weigh on near-term returns, Goldman Sachs believes that attractive ROIC levels from this investment cycle will materialize in medium- to long-term financial statements. In other words, near-term pressure is a natural consequence of large-scale upfront investment cycles and does not indicate a structural flaw in the AI business model.
Broader Market Dynamics
In previous reports on token economics, consumer AI development, and enterprise AI adoption, Goldman Sachs has argued that:
- Market share shifts and declining token pricing will actually boost AI penetration, expanding compute use cases across both consumer and enterprise segments.
- Current market focus is on the evolution of frontier foundation model providers—including market share dynamics and the pricing power of compute relative to open-source models.
- However, in Goldman's view, the cost-performance landscape of compute will ultimately determine the scale and boundaries of the AI economy.
A diverse range of models and agents at varying capability levels, arrayed along a cost-performance curve, will unlock use cases from commoditized, affordable intelligence (with greater price deflation pressure) to frontier, high-end intelligence (with stronger pricing resilience). This will, in turn, support current capital commitments from compute infrastructure providers.
Not All Tokens Are Equal
Goldman Sachs emphasizes that not every token carries the same commercial value, and not every capex dollar will yield identical returns. However, given the broad market opportunity over the next 3–5 years, the firm maintains that capital deployed over the next 18 months will, on aggregate, generate solid returns.
Signals from the Communacopia Technology Conference
This view was reinforced at Goldman Sachs' Communacopia Technology Conference, where participating companies conveyed three key signals:
- The industry has shifted from AI experimentation to implementation.
- Returns from efficiency gains and product iteration cycles are accelerating.
- Corporate focus is moving from maximizing token consumption to optimizing token input-output efficiency, improving enterprise-level returns. This shift signals further upside in enterprise AI adoption, which will directly boost cloud provider revenue and incremental operating profit tied to prior capex.
Consumer AI: A Paradigm Shift
Recent launches of consumer-facing agentic AI products—most notably Meta's Muse—signal a paradigm shift in consumer AI: from conversational interfaces to agent-driven, action-oriented products. The scaling of such agent platforms (with other major players likely to follow with consumer AI platform layers) will drive medium- to long-term compute demand. Supporting monetization models—including subscriptions, advertising, and e-commerce—are expected to roll out gradually. As the boundaries between these business models continue to blur, they will become a core driver of future revenue growth.
Source
智通财经网Western
Part of this Story
Goldman Sachs: US hyperscalers need $1.42 trillion revenue by 2030 for AI ROI