Nvidia (NVDA) Stock Rises Toward Record High as Google Unveils New AI Chips

TLDR

• Nvidia stock rose 0.3% in premarket trading Tuesday, reaching $202.74, closing in on its record close of $207 set in October.
• Google is preparing to unveil a new generation of tensor processing units (TPUs) at Google Cloud Next in Las Vegas this week, developed with Marvell Technology.
• Google’s new TPU push is aimed at inference workloads — the phase where AI models respond to queries — rather than model training, where Nvidia still leads.
• KeyBanc analyst John Vinh kept an Overweight rating on Nvidia with a $275 price target, citing its CUDA software stack as a key competitive barrier.
• Meta signed a multibillion-dollar deal for Google TPUs, and Anthropic expanded its TPU access to up to 1 million chips, but supply constraints remain a challenge for Google.

Nvidia has been on a roll. The chip maker’s stock has climbed 15% over the past month and is now within reach of its all-time closing high. That momentum held Tuesday morning even as Google prepared to make moves in the AI chip space.

NVIDIA Corporation, NVDA

At $202.74 in premarket trading, Nvidia was up 0.3%. The stock is inching toward its record closing high of just over $207, hit back in October 2025.

The gains came as investors looked ahead to upcoming earnings reports from major tech companies. Confidence in Nvidia’s business appears to be building.

The backdrop isn’t entirely clear skies, though. Google is expected to announce a new generation of its tensor processing units — TPUs — at the Google Cloud Next conference in Las Vegas this week.

Google’s Inference Push

According to Bloomberg, Google developed its latest chips in partnership with Marvell Technology. The new chips are focused on AI inference: the stage where a trained model actually answers user queries.

Google has been building toward this for years. Its TPU program now counts Meta as a major customer — the social media giant signed a multibillion-dollar deal to procure TPUs via Google Cloud. Anthropic also expanded its TPU access to up to 1 million chips.

There’s also a structural edge at play. No other leading AI developer manufactures its own chips at comparable volume to Google, tightening the feedback loop between the teams building models and those designing the silicon they run on.

Google has also been opening up its TPU ecosystem. PyTorch users can now access TPUs, and the company has reportedly been piloting on-premises TPU deployments for enterprise customers — a shift away from its historically cloud-only model.

Nvidia’s CUDA Moat

Wall Street isn’t sweating it just yet. KeyBanc analyst John Vinh maintained his Overweight rating on Nvidia Monday with a $275 price target, arguing that the CUDA software stack creates high barriers for any would-be competitor.

Nvidia CEO Jensen Huang has previously said his chips can handle applications “you can’t do with TPUs.” Notably, Google itself still uses Nvidia GPUs alongside its own TPUs for AI projects.

Nvidia’s next-generation Vera Rubin hardware is still expected to be the most advanced AI product on the market when it arrives.

Supply may also be a hurdle for Google’s ambitions. An unnamed startup executive told Bloomberg that TPU scarcity was a real obstacle, with limited access to chips outside of what Google was directing toward “the more elite teams.”

The post Nvidia (NVDA) Stock Rises Toward Record High as Google Unveils New AI Chips appeared first on CoinCentral.