Nvidia and OpenAI Deepen AI Partnership With $100 Billion Deal
Nvidia (NASDAQ: NVDA) and OpenAI remain at the center of the global AI race. Earlier this week, the two companies signed a letter of intent under which OpenAI will deploy at least 10 gigawatts (GW) of Nvidia systems to power training and inference for its next-generation AI models.
![]() |
Nvidia backs OpenAI with record $100B AI investment |
This partnership underscores how deeply Nvidia’s hardware and financial backing are tied to OpenAI’s future. The deal is among the largest infrastructure commitments in AI history.
Nvidia’s $100 Billion Investment in OpenAI
Nvidia is not only providing GPUs and systems but also committing up to $100 billion in direct investment. The capital will be released in stages, linked to the deployment of new AI compute capacity.
According to D.A. Davidson analyst Gil Luria, ranked among the top 5% of Wall Street stock experts, training today’s frontier AI models already requires around 1 GW of compute. By the end of the decade, that number could rise to 5 GW per model, highlighting why OpenAI is aggressively securing capacity.
“OpenAI is locking in as much compute as physically possible. By 2030, these levels—or greater—will be essential for training frontier AI models and handling massive inference demand,” said Luria.
The Scale of AI Compute
Deploying 1 GW of compute in 2026 would equal:
- ~1,667 Nvidia VR200 NVL144 systems
- ~240,000 GPUs
- ~6 zettaflops of FP4 compute
- Equivalent to ~430 million tokens per second for inference
Luria estimates the announced 10 GW buildout will bring over 2 million GPUs online, though the timeline for the remaining 9 GW has not yet been finalized.
Nvidia as “Investor of Last Resort”
Despite the growth potential, Luria warns Nvidia may be shouldering too much of OpenAI’s financial burden:
- Nvidia previously supported CoreWeave’s IPO, which tripled in value.
- OpenAI, however, continues to overcommit resources—reportedly double or triple-booking compute capacity.
- Microsoft holds the right of first refusal on OpenAI’s compute but has allowed OpenAI to commit:
- $300B with Oracle
- $100B for backup servers from cloud providers
- $25B to CoreWeave
- Unquantified billions in Europe and the Gulf
Meanwhile, OpenAI is still losing tens of billions annually on ~$12 billion in revenue. Even under optimistic projections, it expects only $200 billion revenue by 2030.
Luria cautions that Nvidia risks becoming the “investor of last resort” if traditional investors don’t share more of the funding burden.
Nvidia Stock Forecast
Despite these concerns, Luria remains bullish on Nvidia stock:
![]() |
Nvidia stock market chart with upward trend. |
- Reaffirmed Buy rating
- $210 price target (~19% upside)
- Wall Street consensus: Strong Buy
- 37 Buy ratings
- 2 Holds
- 1 Sell
- Average 12-month target: $212 (~20% upside)
AI Investment Surge
- Nvidia and OpenAI are doubling down on AI infrastructure.
- $100B investment marks Nvidia as more than a supplier—it’s now a financial backer.
- AI compute demand could reach unprecedented levels by 2030.
- Risks remain as OpenAI overcommits resources, but Nvidia continues to attract Wall Street confidence.
Comments
Post a Comment