Nvidia’s reported Perplexity talks show AI finance moving beyond chips and data centers

Nvidia is reportedly in talks to invest in Perplexity as part of an equity round that would value the AI search and agent startup at more than $30 billion, adding another example of the chipmaker’s expanding role in the AI economy. Reuters reported on August 23, citing The Information, that the discussion involves a valuation above $30 billion, while noting that Reuters could not immediately verify the report. The talks have not been announced as a completed transaction, but the possibility is notable because Nvidia is no longer only selling the processors that power AI systems. It is increasingly shaping the financial and software ecosystem around those systems.

Perplexity began as an AI-native search engine and has since moved into broader productivity tools, including agentic software that can help professionals automate computer tasks. The Information reported that Perplexity’s annualized revenue has climbed sharply this year and that the company has developed closer ties with Nvidia. Those ties matter because Perplexity depends on a mix of open and closed models, cloud infrastructure and efficient inference to deliver answers quickly. If Nvidia becomes a larger investor, the relationship could strengthen Perplexity’s access to hardware strategy, developer tooling and enterprise credibility.

For Nvidia, the reported talks fit a pattern. The company has invested in cloud providers, data-center builders, model developers and AI software firms that either buy its chips or expand demand for accelerated computing. It has also pushed its own open model work through the Nemotron family and industry collaborations. A potential investment in Perplexity would sit at the intersection of those priorities: search, agents, inference, enterprise usage and open model routing.

The move would come shortly after Nvidia’s reported Poolside arrangement, in which the chipmaker agreed to license AI model development technology and recruit more than 100 employees while also investing in the company. That deal raised questions about whether Nvidia is using its balance sheet to acquire talent and software capabilities without buying entire startups outright. Perplexity would be different because the report describes an equity investment rather than a confirmed licensing-and-hiring package, but the strategic logic overlaps. Nvidia wants more influence over the software layers that make its hardware indispensable.

The possible valuation also says something about investor appetite for AI applications that show usage beyond novelty. Perplexity competes in a difficult space where search, chatbots, browser agents and workplace assistants are converging. Google, OpenAI, Anthropic and xAI all have reasons to push into similar workflows. Perplexity’s challenge is to remain differentiated while paying for expensive inference and persuading businesses that its answers and agents are reliable enough for daily work. A large Nvidia-backed financing round could buy time, capacity and confidence as that market develops.

There are risks on both sides. Nvidia’s growing web of investments has drawn scrutiny because many recipients are also customers or infrastructure partners, raising questions about circular demand in the AI boom. Perplexity faces intense competition and must keep improving product quality while managing content licensing, attribution and enterprise trust. A reported valuation above $30 billion gives the company prestige, but it also raises the bar for future revenue growth and eventual public-market expectations.

Still, the talks illustrate a broader shift in AI finance. The most powerful chip supplier in the market is becoming a capital allocator, ecosystem architect and sometimes a software strategist. If the Perplexity investment proceeds, it would reinforce the idea that control of AI’s future will not be determined by hardware alone. The winners will also be decided by which platforms can turn compute into useful daily workflows and which investors can afford to keep those platforms running while the market matures.