According to U.S. tech media outlet The Information, Nvidia is in talks to participate in a new equity funding round for AI search company Perplexity, which would value the company at over $30 billion. Reuters confirmed the report. If the deal is completed, Perplexity's valuation would increase by more than 50% from approximately $20 billion after its previous funding round about a year ago. Nvidia was already an investor in Perplexity, and these talks represent a deepening of that existing equity relationship. Neither party has commented on the matter.
Underpinning this valuation jump is Perplexity's near-vertical revenue growth curve. According to The Information, citing sources familiar with the matter, Perplexity's annual recurring revenue (ARR) has grown from under $250 million as of January 2026 to over $750 million by August—tripling in eight months. The core product driving this growth is a cloud-based AI agent tool called "Perplexity Computer." The product allows professional users to automate computing tasks, and its pricing model has shifted from traditional advertising to a usage-based subscription, significantly boosting revenue per user.
What Is Perplexity, and Why Are Its Ties with Nvidia So Deep
To understand this potential deal, one must first understand Perplexity's product positioning. Perplexity was founded in San Francisco in August 2022 by four co-founders, including Aravind Srinivas, who previously worked at OpenAI and Google DeepMind, and Denis Yarats, who previously worked at Microsoft Bing and Quora. Its core product is an "answer engine": rather than returning lists of links like traditional search engines, it combines large language models with real-time data retrieval to deliver answers with traceable sources directly to users. This architecture places Perplexity at the intersection of both the search and AI agent markets.
Perplexity has raised more than $1.7 billion in cumulative funding to date, with investors including New Enterprise Associates (NEA), Accel, SoftBank Group, and Amazon founder Jeff Bezos. Nvidia has been an investor across multiple early rounds.
Why Nvidia Is Doing This: From Selling Shovels to Buying into the Gold Mine
Nvidia dominates the AI training and inference chip market, but this dominance carries an inherent vulnerability: it depends on sustained strong downstream demand for computing power. Concerns have already surfaced in the market about whether Nvidia's GPU shipments can maintain rapid growth as major cloud computing vendors slow their capital expenditure growth. Against this backdrop, Nvidia's strategic choice is to extend toward the application layer of the value chain. By taking equity stakes in high-traffic AI application gateways, it ensures its chip demand comes not only from a few large cloud vendors but is embedded in a broader range of enterprise and individual workflows.
In this logic, Perplexity's role is that of a traffic gateway connecting end users with underlying models. The strategic value of such gateways lies in the fact that regardless of which company's large model runs underneath, user-side traffic must pass through Perplexity's interface—and Perplexity's infrastructure requires substantial GPU and CPU computing power. By investing in Perplexity, Nvidia is essentially locking in a high-growth customer for future computing power procurement contracts in advance.
Vera CPU: Technology Ties Precede Capital Ties
What makes this investment negotiation more compelling is the technology-layer collaboration that has already been put in place. In July 2026, Perplexity publicly announced it would adopt Nvidia's newly released Vera CPU to run its AI agent workloads. The Vera CPU is the processor core of Nvidia's Vera Rubin platform, designed on the ARM architecture with 88 Olympus cores and up to 1.2 TB/s of LPDDR5X memory bandwidth. Nvidia positions it as "the first CPU built for AI agents."
Nate Kupp, Perplexity's Vice President of Enterprise Infrastructure, disclosed in the announcement that the Vera CPU processes agent programming tasks approximately 1.5 times faster than traditional server CPUs. This means Perplexity has already tested and adopted Nvidia's new product in real workloads. The technology tie was completed before the capital tie.
Circular Financing Concerns and Valuation Debate
The deal also faces critical perspectives from the market. Some analysts describe Nvidia's investments in AI application-layer startups as "circular financing": Nvidia injects capital into Perplexity, and Perplexity then uses that money to purchase Nvidia GPUs and CPUs, with funds effectively circulating in a closed loop between Nvidia's balance sheet and its portfolio companies. Whether this structure raises related-party transaction concerns at the accounting and regulatory level is one of the main questions raised externally.
There is also a more fundamental valuation question. Perplexity's ARR grew from $250 million to $750 million in eight months. Based on static revenue multiples, a $30 billion valuation corresponds to a price-to-sales ratio of approximately 40x. This multiple is not uncommon in the high-growth AI sector, but it rests on a premise: that Perplexity's revenue growth rate can be sustained. Whether the enterprise subscription growth driven by Perplexity Computer is replicable, and whether the usage-based pricing model is resilient during an economic downturn, remain open variables.
Industry Implications for All Parties
For Google, Perplexity is one of the most significant search challengers today. Nvidia's capital backing will strengthen Perplexity's staying power in the search advertising replacement market, but the competitive landscape between the two will not fundamentally change in the short term because of a single funding round—Google's advantages in search remain built on more than a decade of index accumulation, distribution channels, and advertising systems.
For developers and enterprise users, the signal from these negotiations is that Perplexity's viability now has greater assurance, and the risk of betting on Perplexity as the gateway to enterprise AI workflows is decreasing. Microsoft previously signed a $750 million Azure cloud services agreement with Perplexity, and the major cloud vendors' positioning around Perplexity has already formed a competitive landscape. Nvidia's entry will make that landscape even more complex.
For Nvidia's competitors, especially AMD and Intel, this investment means Nvidia is turning its chip customers into equity-bound ecosystem partners. Once this model is replicated at the application layer, competitors seeking to take share from Nvidia will need not only to catch up in chip performance but also to break through the ecosystem lock-in created by this dual binding of capital and technology.
Strategic Outlook: What Signals to Watch Next
The first signal to watch is the specific structure of the investment terms. If Nvidia ultimately completes this investment in the form of equity in exchange for computing power commitments (i.e., Perplexity commits to procuring a certain scale of Nvidia computing resources over several years), the "circular financing" criticism will be harder to refute, and regulators may turn their attention sooner. Conversely, if it is a pure cash equity investment, there is relatively less room for controversy.
The second signal is Perplexity's IPO timeline. In a June 2026 interview with CNBC, Perplexity CEO Aravind Srinivas said the company plans to go public in 2028, independent of market reactions to Anthropic's and OpenAI's listings. Between the $30 billion private valuation and the 2028 listing target, the value-creation space left for existing investors will depend on the market environment at that time—which is also the basic return logic Nvidia needs to calculate in these negotiations.
The third signal is the enterprise adoption rate of Perplexity Computer. The entire ARR growth story centers on this agent product. If it can push ARR from $450 million (March data) to $750 million within six months, then its paid customer stickiness and expansion trajectory will become the core basis for the next round of valuation discussions. What the market can currently see is the result numbers, but the customer composition and churn rate driving them have not yet been publicly disclosed.
From a broader perspective, the evolving relationship between Nvidia and Perplexity confirms a structural change taking place in the AI industry: the position of a pure hardware supplier is increasingly unable to command high premiums in the AI value chain. Computing power companies need to extend into the application layer to secure their place in the distribution of AI's returns. This is not a choice Nvidia faces alone—it is a shared challenge for the entire computing power industry.
© 2026 Winzheng.com 赢政天下 | 转载请注明来源并附原文链接