Nvidia (NVDA) shares fell toward $225 on October 12, pulling back from the record high of $243.37 reached on October 6. The stock weakened alongside the broader chipmaking sector as investors weighed the rising cost of building out artificial intelligence infrastructure.

The retreat followed Microsoft's announcement of a new custom processor, renewed questions about OpenAI's revenue figures, and reports of heavy borrowing to fund AI hardware purchases amid climbing data center costs. Nvidia remains a leading supplier of data center chips, although several of its largest customers are developing of their own.

Microsoft's Maia 300 Targets AI Costs

Microsoft expects its Maia 300 chip to produce 40% more tokens per megawatt than leading GPU systems, according to the company. A rack equipped with 96 of the accelerators could deliver up to seven times the token output of the Maia 200, Microsoft said. Efficiency claims carry weight because electricity and cooling have become the practical limits on how quickly AI data centers can scale.

The company plans to begin production deployments in 2027 across its Azure cloud services. Microsoft's in-house processors could reduce its dependence on outside suppliers, although Nvidia GPUs remain widely used for AI training and other computing workloads. Microsoft is not alone in that push: Google's Tensor Processing Units and Amazon's Trainium chips reflect the same effort by Nvidia's largest cloud customers to design their own AI silicon.

OpenAI Revenue Figures Draw Scrutiny

OpenAI's annualized revenue reached roughly $50 billion in September, Reuters reported, coming in below earlier estimates near $68 billion. The news agency attributed part of the gap to how sales made through cloud partners are counted in the calculations. Annualized revenue projects a company's current rate over twelve months.

The revenue debate added to concerns about concentration in AI-linked stocks. One analysis found that 42 AI companies generated 67% of the S&P 500's price gains since January 2024. Investors are watching closely whether massive infrastructure spending ultimately translates into revenue. Because OpenAI ranks among the largest buyers of AI computing capacity, its revenue trajectory is widely treated as a gauge of demand for data center chips.

Nvidia Weighs Deeper Bet on Reflection AI

Nvidia is considering acquiring artificial intelligence startup Reflection AI or raising its existing investment, the Financial Times reported. Nvidia has already invested roughly $800 million in the company, which develops open-weight AI models. Talks remain preliminary, and no formal deal has been announced.

An acquisition would broaden Nvidia's business beyond processors and into AI software, while adding another investment commitment to its portfolio. Software already anchors Nvidia's ecosystem through CUDA, the programming platform developers have used for nearly two decades to run workloads on its GPUs. Semiconductor stocks have declined in tandem as traders assess customers' spending plans and financing arrangements.

Debt-Funded AI Spending Keeps NVDA in Focus

SpaceX is reportedly seeking $40 billion in financing to purchase Nvidia chips through a combination of loans and bonds, Reuters reported. The plan underscores how AI customers increasingly rely on borrowing to finance expensive chips and data centers. No financing agreement has been announced. Debt obligations persist regardless of how AI demand evolves, one reason investors now track financing arrangements as closely as order books.

NVDA stock traded near $226 on Monday after dipping below $225 during another losing session. The October peak remains a reference point for traders, who are monitoring AI orders, infrastructure spending, and how quickly customers deploy competing processors.

Source: Blockonomi