Nvidia eyes Perplexity AI investment at $30 billion valuation

This move functions as a strategic insurance policy to keep the AI search compute layer tethered to Nvidia silicon. By deepening its stake in Perplexity, Nvidia proactively blocks rival chipmakers like AMD and Cerebras from capturing high-frequency inference workloads within the search ecosystem. This reflects a broader pattern where Nvidia and Wall Street mobilize $500 billion for AI compute financing to create a self-reinforcing demand cycle for its chips. As Perplexity integrates with 800 million Samsung Bixby devices, Nvidia secures a massive distribution channel for its architecture. Watch for Perplexity's progress toward a 2028 initial public offering as a benchmark for Nvidia's return on investment.

Perplexity AI has rapidly expanded its infrastructure partnerships as it scales its AI search platform. In early 2026, Perplexity announced a multi-year deal with Microsoft Azure to handle a significant portion of its inference workloads, a move that diversified its compute dependencies beyond any single hardware vendor. The company also launched Perplexity Computer, its agentic AI product, in May 2026, which automates multi-step tasks across applications and significantly increases per-query compute requirements. That product launch intensified demand for GPU capacity and made Perplexity's infrastructure alignment a strategic priority for chipmakers competing for inference market share. On the business side, Perplexity's valuation trajectory has attracted a broad coalition of investors beyond Nvidia. SoftBank led a $500 million funding round in early 2026 that valued Perplexity at $18 billion, and Jeff Bezos personally participated in that round through his family office. The company's annualized revenue run rate reached $750 million by mid-2026, driven largely by enterprise subscriptions and API licensing. Perplexity also signed a distribution agreement with Samsung in April 2026 to integrate its search capabilities into Bixby across 800 million devices, creating a consumer distribution channel that dramatically increases query volume and, by extension, inference compute demand. This commercial momentum is what makes the Nvidia investment strategically distinct from a pure financial bet. The competitive dynamics around Perplexity's compute stack reflect a broader industry pattern. Nvidia and AWS lead shift toward hybrid AI architectures for inference in June 2026, though Perplexity was not among them. Cerebras Systems separately disclosed a partnership with an unnamed AI search provider in July 2026, which industry analysts speculated could target Perplexity's competitors. Nvidia's interest in Perplexity thus represents a defensive posture: by securing equity alignment with the fastest-growing AI search company, Nvidia ensures that its H200 and successor B300 GPUs remain the default inference substrate for a platform processing billions of queries monthly. The investment also positions Nvidia to benefit from Perplexity's anticipated 2028 IPO, which would provide liquidity and validate the compute-demand thesis.