OpenAI CFO Confirms ROI-Driven Spending for Chip Makers

OpenAI’s Chief Financial Officer Sarah Friar confirmed during a recent interview that the company will continue prioritizing strong returns on investment for its AI infrastructure buildout. This commitment to ROI-based decision-making provides a clear operational framework for the six major suppliers currently financing the sector’s expansion. The assurance that spending is driven by tangible business returns rather than speculative growth stabilizes the revenue pipeline for hardware providers.

The direct beneficiaries of this sustained capital expenditure are the semiconductor and infrastructure firms already integrated into OpenAI’s supply chain. Nvidia and AMD lead the list, with their data center revenues surging due to multi-gigawatt scale deployments. The financial link between OpenAI’s compute commitments and these manufacturers’ quarterly results is now a primary driver of stock performance in the technology sector.

Nvidia and AMD Capture Core Compute Demand

Nvidia reported second-quarter revenue of $96.22 billion, a 105.8% year-over-year increase, with data center sales accounting for $89.02 billion of that total. CFO Colette Kress noted that OpenAI’s existing and planned commitments represent approximately 12 gigawatts of Nvidia compute capacity, anchored by a 4.25-gigawatt AI factory campus in Portsmouth. Nvidia’s third-quarter guidance stands at $108 billion, reflecting continued demand for its accelerated computing hardware.

AMD posted second-quarter revenue of $11.54 billion, with data center revenue rising 107% year-over-year to $6.72 billion. CEO Lisa Su highlighted multi-generation gigawatt-scale deployments with both OpenAI and Meta. Additionally, AMD announced a strategic partnership with Anthropic to deploy up to two gigawatts of MI450 GPUs. These contracts underscore the growing market share AMD is gaining in the high-performance AI accelerator space.

Custom Silicon and Infrastructure Support Revenue

Broadcom reported fiscal third-quarter revenue of $29.59 billion, driven by AI chip revenue of $16.70 billion, a 221% increase. CEO Hock Tan confirmed that OpenAI’s Jalapeno accelerator is on track for a 1.3-gigawatt deployment in 2027. Tan also disclosed visibility into $230 billion in fiscal 2028 AI semiconductor revenue, linking each gigawatt of deployed compute to approximately $30 billion in annual recurring revenue.

Supporting infrastructure providers are also seeing significant growth. Arista Networks reported second-quarter revenue of $3.04 billion, up 37.7%, with AI fabric momentum exceeding 100 cumulative customers. Vertiv Holdings raised its full-year guidance to net sales between $13.80 billion and $14.20 billion. Eaton’s CEO noted that total US data center backlog has grown to 307 gigawatts, representing 15 years of work at current build rates.

Advertising Revenue Supports Compute Expenditures

Sarah Friar explained that OpenAI’s advertising business reached $1 billion in revenue within seven months, making it the fastest-growing ad platform in history. By reducing model pricing by 80%, the company increased demand tenfold, securing the top-ranked position on OpenRouter. This strategy converts the roughly 90% of ChatGPT users on the free tier into a revenue source that helps service the substantial capital expenditure bill.

According to GN auto stocks/technology: chip stocks, the financial logic remains that every new dollar of compute must generate a measurable return. Friar’s comments dismissed concerns about a safety-driven slowdown, emphasizing that investment decisions will continue to be grounded in strict ROI metrics. This approach ensures that the massive infrastructure buildout remains aligned with sustainable business fundamentals rather than speculative hype.