Key Takeaways
Nvidia revealed details about its next-generation AI processor at its annual developer conference on Tuesday, March 18.
However, the announcement that Vera Rubin GPUs will ship in 2026 failed to invigorate NVDA, which has struggled since the release of DeepSeek R1, which sparked concerns about ongoing demand for AI chips.
Since the mid-2010s, Nvidia’s high-performance data center GPUs have represented the gold standard for industrial AI development.
The firm’s current top-of-the-range Blackwell solutions hit the market in late 2024 and are expected to drive demand for the foreseeable future.
However, Nvidia has already started developing Blackwell’s successor.
On the first day of the firm’s GTC 2025 conference, Huang confirmed that the first Rubin solutions would hit the market in early 2026, with an “Ultra” version to follow the year after.
In the past, Nvidia’s AI product releases have catalyzed strong stock market gains in the following weeks and months.
However, in the hours after Huang’s Rubin announcement, NVDA fell 3.4% amid a wider selloff of U.S. equities.
Although negative market sentiment has affected all major American tech stocks in the past month, fears over DeepSeek’s long-term impact have exacerbated Nvidia’s stock market performance.
The release of DeepSeek R1 in January 2025 triggered concerns that DeepSeek’s more efficient AI model could dampen demand for Nvidia GPU.
After the release, Nvidia’s market cap plummeted by nearly $600 billion, marking the stock’s worst performance on record.
NVDA hit a six-month low on March 10, and despite tentative signs of recovery in the following days, it currently trades roughly 20% below its pre-crash level.