NextFin News - In a week defined by high-stakes earnings reports from the world’s largest technology conglomerates, a curious divergence has emerged in the semiconductor and cloud computing sectors. On February 5, 2026, Amazon.com Inc. reported its fourth-quarter 2025 financial results, headlined by a staggering commitment to invest approximately $200 billion in capital expenditures (capex) throughout 2026. This figure, primarily directed toward artificial intelligence (AI) infrastructure and the expansion of Amazon Web Services (AWS), represents one of the largest single-year investment cycles in corporate history. However, the anticipated "halo effect" for Nvidia Corporation, the primary provider of AI accelerators, has failed to materialize. As of Friday, February 6, 2026, Nvidia’s stock remains sluggish, continuing a period of consolidation that has frustrated investors expecting a breakout on the back of Big Tech’s aggressive spending.
According to Amazon’s official earnings release, the company’s AWS segment saw sales increase 24% year-over-year to $35.6 billion in the fourth quarter, its fastest growth in over three years. Chief Executive Officer Andy Jassy emphasized that the $200 billion capex plan is a direct response to "strong demand for existing offerings and seminal opportunities like AI." While such a massive injection of capital into data centers would historically trigger a rally for Nvidia, the market reaction has been muted. Nvidia’s shares have traded with low volatility, mirroring a broader trend seen earlier in the week when Alphabet Inc. announced its own plan to spend between $175 billion and $185 billion in 2026. Despite these combined commitments totaling nearly $400 billion from just two players, Nvidia’s stock has struggled to reclaim its previous highs, signaling a fundamental shift in how the market values the AI hardware leader.
The primary cause for this sluggishness is the accelerating trend of "silicon independence" among Nvidia’s largest customers. During the earnings call, Jassy highlighted the triple-digit growth of Amazon’s custom chip business, noting that its Trainium and Graviton processors now have a combined annual revenue run rate exceeding $10 billion. Specifically, Amazon revealed that its Trainium2 chips are already powering the majority of inference on its Bedrock service and that the newly production-ready Trainium3 is seeing such high demand that supply is expected to be fully committed by mid-2026. This internal shift suggests that while Amazon is spending more than ever, a growing percentage of that capital is being diverted away from Nvidia’s H-series and Blackwell architectures in favor of proprietary, cost-optimized solutions.
Furthermore, the market is currently grappling with a "valuation reset" for the semiconductor sector. According to analysis from Intellectia AI, Nvidia’s forward price-to-earnings (P/E) ratio has settled into a "fair zone" of approximately 29.6, a significant contraction from the exuberant multiples seen in 2024 and early 2025. Investors are no longer rewarding the mere promise of increased capex; they are demanding proof of sustainable margins and long-term utility. The fact that Amazon’s free cash flow decreased to $11.2 billion in 2025—down from $38.2 billion the previous year—due to AI-related property and equipment purchases has also introduced a note of caution. There is a growing concern that the "capex arms race" may eventually lead to overcapacity, or that the return on invested capital (ROIC) for these cloud providers may take longer to materialize than initially projected.
From a technical perspective, Nvidia’s stock is facing a "digestion phase." After the meteoric rises of the past two years, the stock is finding a new equilibrium where the massive revenue beats are already priced in. Even with projections that Nvidia’s revenue could jump another 52% in fiscal 2027, as suggested by some analysts following the Alphabet and Meta spending plans, the immediate upside is capped by the reality of a diversifying supply chain. Competitors like Advanced Micro Devices (AMD) are also making inroads; recent reports indicate AMD is rapidly increasing its market share in the AI chip sector, with some analysts projecting it could eventually capture a significant portion of the non-proprietary accelerator market.
Looking ahead, the trajectory for Nvidia will likely depend on the successful ramp-up of its next-generation Vera Rubin platform. While hyperscalers like Alphabet have identified themselves as early adopters of this new architecture, the window of absolute dominance is narrowing. The 2026 fiscal year appears to be a transition period where the narrative shifts from "scarcity of supply" to "efficiency of spend." For Nvidia to break out of its current doldrums, it will need to demonstrate that its software ecosystem and networking capabilities (such as InfiniBand and Spectrum-X) provide a performance moat that custom silicon cannot easily replicate. Until then, the stock is likely to remain sensitive to the broader macroeconomic environment and the specific ROIC metrics reported by its "Magnificent Seven" customers, rather than just the raw dollar amount of their infrastructure budgets.
Explore more exclusive insights at nextfin.ai.
