Remember that massive $100 billion deal between Nvidia Corp. (NVDA) and OpenAI that generated so much buzz? Turns out it's not actually done yet. Like, not even close to done.

Still Working Out The Details

Speaking at the UBS Global Technology and AI Conference in Scottsdale on Tuesday, Nvidia's CFO Colette Kress dropped some reality on investors: the much-hyped OpenAI partnership is still sitting at the letter-of-intent stage, according to Fortune. No "definitive agreement" has been signed, she confirmed.

Before you panic, Kress was quick to stress that Nvidia's relationship with OpenAI remains strong. More importantly, the company's current sales forecasts aren't banking on this megadeal actually closing. So Nvidia isn't sitting around waiting for OpenAI's signature before it can pay the bills.

For now, OpenAI is doing what it's been doing all along: buying computing power through cloud partners like Microsoft (MSFT) and Oracle (ORCL) instead of going direct to Nvidia as the proposed deal would allow. Kress acknowledged that OpenAI "does want to go direct," but they're still hammering out the definitive agreement. These things take time, apparently.

OpenAI's Rocky Road

The uncertainty around this deal isn't exactly new. Nvidia (NVDA) flagged it in its latest earnings report, though CEO Jensen Huang remained optimistic during the earnings call, calling OpenAI a "once in a generation company." That's high praise, even if the paperwork isn't finished.

Not everyone shares Huang's enthusiasm. Commentator Jim Cramer has suggested OpenAI's future might be shakier than it appears, particularly with Alphabet Inc. (GOOGL) (GOOG) making waves with its Gemini 3 AI model. The competitive pressure is real.

Cramer later elaborated on CNBC that OpenAI's biggest problem isn't just competition from Google but its financial situation. Unlike the tech giants that can borrow money at favorable rates, OpenAI is already carrying significant debt. Cramer suggested two potential fixes: settle that ongoing New York Times lawsuit to reduce legal expenses, or get Microsoft to increase its stake and shore up the balance sheet.

The Google Threat That Isn't

Meanwhile, Nvidia's (NVDA) dominance in the AI chip market has faced scrutiny lately. Alphabet has emerged as a serious competitor with its focus on Tensor Processing Units (TPUs), and the stock performance tells an interesting story. Over the past month, Alphabet stock jumped 11.23%, while Nvidia stock dropped 12.29%, according to market data.

On Tuesday, Nvidia stock closed up 0.86% at $181.46, so it's not all doom and gloom.

When asked about competitive pressure from other chipmakers, Kress didn't mince words: "Absolutely not" bothered, she said. Her argument? Nvidia's (NVDA) real advantage isn't any single chip but its entire ecosystem. That includes the hardware, CUDA software platform, and an expanding catalog of specialized software tools that keep customers locked into the Nvidia universe.

It's the classic platform play: get customers invested in your full stack, and switching becomes exponentially harder even if competitors offer cheaper individual components. Kress emphasized that Nvidia's focus is on supporting various model builders and enterprises with comprehensive solutions, not just selling chips.

The strategy seems to be working, at least according to certain metrics. Nvidia scores high on Momentum, Growth, and Quality factors, with a favorable long-term price trend, though the short and medium-term trends have been weaker.

So where does this leave us? The OpenAI deal is still in limbo, Google is nipping at Nvidia's heels, and OpenAI itself faces both competitive and financial headwinds. But Nvidia's CFO is projecting confidence, betting that the company's ecosystem advantage will outlast any near-term turbulence. Whether that confidence is justified depends on whether customers agree that Nvidia's full-stack approach is worth the premium over alternatives.