Oracle stated on social media platform X Monday that it remains "highly confident" in OpenAI's funding capacity, as the cloud provider sought to address investor concerns amid reports of a stalled $100 billion Nvidia investment and chip performance issues.
The cloud provider's stock fell 2.79% to $160.06 despite the public vote of confidence. Oracle faces investor anxiety over its own $50 billion 2026 fundraising plan for AI infrastructure expansion and approximately $100 billion debt load.
"The NVIDIA-OpenAI deal has zero impact on our financial relationship with OpenAI," Oracle stated on X February 2. "We remain highly confident in OpenAI's ability to raise funds and meet its commitments."
Nvidia's planned $100 billion investment in OpenAI, first announced in September 2025, remains unclosed months later. The chipmaker's CEO Jensen Huang called the figure "never a commitment" last week, though he confirmed Nvidia will participate in OpenAI's current funding round.
OpenAI has reportedly grown dissatisfied with Nvidia's latest AI chips for inference workloads, seeking alternatives since last year. Eight sources told Reuters the ChatGPT maker needs hardware covering roughly 10% of future inference computing needs.
Inference, the stage where trained AI models generate responses to user queries, requires different architecture than training. OpenAI specifically needs faster processing for software development tools and AI-to-AI communication.
The performance issues became visible in OpenAI's Codex coding product, where staff attributed limitations to Nvidia's GPU-based hardware. CEO Sam Altman acknowledged January 30 that coding customers "will put a big premium on speed for coding work."
OpenAI explored partnerships with Cerebras, Groq, and AMD for chips with large amounts of embedded SRAM memory. This architecture offers speed advantages for real-time applications by reducing data retrieval delays.
Nvidia responded with defensive moves, including a $20 billion licensing deal with Groq that reportedly shut down OpenAI's negotiations. The chipmaker also hired away Groq's chip designers while describing the technology as "highly complementary" to its roadmap.
OpenAI secured a commercial agreement with Cerebras last month to address inference speed concerns. Cerebras specializes in wafer-scale chips with substantial on-chip memory.
Despite the behind-the-scenes friction, both companies maintain public harmony. Altman posted on X Tuesday that "We love working with NVIDIA and they make the best AI chips in the world. We hope to be a gigantic customer for a very long time."
Huang dismissed tension reports as "nonsense" Saturday, reiterating plans for significant OpenAI investment. Nvidia stated customers choose its chips because they "deliver the best performance and total cost of ownership at scale."
The developments come against Oracle's massive September 2025 deal with OpenAI, which commits the ChatGPT maker to purchase $300 billion worth of compute power over five years. OpenAI has committed to approximately $1.4 trillion in infrastructure spending despite 2025 annualized revenue of just over $20 billion.
Microsoft and Nvidia shares also faced pressure as investors assess exposure to OpenAI through commercial partnerships. Oracle's five-year credit default swaps fell 17%, indicating some investor confidence in the company's debt management efforts.
The AI hardware market is shifting as inference gains importance relative to training. McKinsey projects the inference market could be 2-5 times larger than training, reaching over $200 billion annually by 2028.
Competitors like Anthropic's Claude and Google's Gemini already use Google's custom tensor processing units optimized for inference. These TPUs offer performance advantages over general-purpose GPUs for certain workloads.
Oracle's statement reflects broader industry concerns about AI infrastructure economics. With much capacity built on demand and OpenAI committing trillions in spending, investors worry about exposure if demand softens.
The cloud provider plans to raise up to $50 billion this year using debt and equity financing for additional data center capacity. Its shares initially rose 2% Monday on the fundraising announcement before declining.
Nvidia stock fell 2.9% February 2 amid the conflicting signals, though it rebounded nearly 1% in premarket trading Tuesday. Stocktwits sentiment shifted to "bearish" from "neutral" as uncertainty persists.
OpenAI's computing infrastructure leader Sachin Katti emphasized the depth of the Nvidia relationship on X, calling it "deep, ongoing co-design" rather than a vendor relationship. He noted OpenAI's compute capacity would accelerate from roughly 1.9 GW in 2025.
The situation reveals strategic divergence as AI applications mature. While Nvidia dominates training, inference is becoming a separate battleground where specialized hardware gains importance.
Oracle's public confidence statement attempts to stabilize its position amid these industry shifts. The cloud provider's fortunes remain tied to OpenAI's ability to fund its ambitious infrastructure commitments while navigating chip supplier relationships.
