NVIDIA is transforming from a chip manufacturer into a central capital force within the entire artificial intelligence industry. Over the past 16 months, NVIDIA has committed approximately $90 billion to investment and partnership deals, covering more than 145 companies. This portfolio spans AI model developers, cloud computing service providers, and infrastructure suppliers. This scale is now comparable to the largest venture capital arms of major tech firms, positioning NVIDIA as one of the most aggressive dealmakers in the technology sector. The company is set to report quarterly earnings, which the market views as a key indicator for global AI spending. Concurrently, NVIDIA announced it will establish its first research center in Singapore, marking its second research facility in the Asia-Pacific region. This center will focus on embodied AI and improving AI infrastructure efficiency. The move aligns with Singapore's strategy to position itself as a regional AI hub, further highlighting NVIDIA's ambition to expand its global AI footprint. This massive deal-making campaign accelerates the expansion of the AI ecosystem while also placing NVIDIA at the intersection of clients, suppliers, and potential competitors. This activity has already drawn scrutiny from global regulators. Where does the $90 billion come from? Two parallel funding streams. According to company filings and data, NVIDIA committed roughly $47 billion for investments and partnerships in the fiscal year ending January 25, followed by an additional $43 billion in the subsequent four months. This expenditure represents about 40% of NVIDIA's operating cash flow in its latest fiscal year, far exceeding the approximately 6% cash flow investment ratio of Alphabet—traditionally seen as one of the most active large-tech investors in startups. Two parallel deal-making mechanisms operate within NVIDIA. The company has a venture capital arm called NVentures, but business development teams have led most recent deal flows, according to sources. Investments often land alongside broader commercial cooperation agreements. A San Francisco lawyer described a process where NVIDIA's technical teams engage with target companies, often followed by business development personnel "arriving with a checkbook," creating "two parallel lines of conversation." Ecosystem Lock-in: From Chip Interconnects to Open-Source Models NVIDIA's investment logic is not purely financial return. Instead, it uses capital as a tether to deeply bind partners to its own technology ecosystem. Patrick Little, CEO of semiconductor design startup SiFive, revealed that NVIDIA invested in SiFive after the two companies reached an agreement to make SiFive's chip designs compatible with NVLink—NVIDIA's proprietary interconnect technology. He stated that NVIDIA's investment rationale is to "ensure the two solutions always work well together." NVIDIA reached a similar arrangement with chip designer Marvell, investing $2 billion in March alongside a cooperation agreement to ensure Marvell's future custom chips are compatible with NVLink. Marvell is the chip designer behind Amazon's Trainium AI accelerators. On the software front, NVIDIA is also actively encouraging portfolio companies to adopt its open-source AI model, Nemotron, according to two sources familiar with the matter. Jensen Huang hopes Nemotron can replicate the success of CUDA—the proprietary software platform that forms one of NVIDIA's strongest competitive moats. A venture capitalist bluntly stated, "Founders are starting to realize that if you build on NVIDIA's ecosystem, you can raise money from Jensen." Cloud Computing Newcomers: A Trinity of Client, Supplier, and Shareholder Within NVIDIA's deal landscape, its bets on a new generation of cloud computing companies are particularly notable and also a source of concern for analysts. Jensen Huang has publicly stated that CoreWeave "would not exist without NVIDIA's support." NVIDIA's support for these emerging AI infrastructure providers is partly because major hyperscale cloud service providers like Google and Amazon are both NVIDIA's largest customers and a growing competitive threat as they develop their own chips. Earlier this month, NVIDIA reached a deal with emerging cloud computing company Iren, committing to pay $3.4 billion over five years to lease its GPU computing power, while also investing up to $2.1 billion for an equity stake. This arrangement places NVIDIA in the triple roles of client, supplier, and potential shareholder. NVIDIA's largest single deal last year was a $20 billion agreement with chip designer Groq, covering technology licensing and talent acquisition. As AI workloads shift towards "inference" computing, where Groq's processors hold a distinct advantage, NVIDIA has already launched products based on Groq's technology. Supply Chain Positioning: Securing Critical Capacity Beyond equity investments, NVIDIA had also committed $95 billion by the end of January to lock in component supply and manufacturing capacity. Recent major supplier deals announced include: a $2 billion investment each in photonics companies Coherent and Lumentum, and a $3.2 billion investment in warrants for optical fiber manufacturer Corning, which produces fiber used in high-speed data centers. All three are significant suppliers to Apple, and NVIDIA's investments are positioning it to become a core client for these suppliers. Moon Surana, a portfolio manager at asset management firm Harding Loevner, noted that NVIDIA's funding supports supplier expansion plans while "enhancing NVIDIA's bargaining power within the supply chain in an environment where capacity remains tight." Regulatory Risks and Singapore Expansion The scale of the deal-making campaign has attracted attention from global regulators. NVIDIA disclosed in its annual report that regulators in the US, EU, UK, and elsewhere have issued "broad requests for information" regarding its "investments, partnerships, and other agreements with foundational model development companies," as well as customer, supplier, and partner agreements. On its global expansion front, NVIDIA's announcement of its first Singapore research center marks a strategic move. The center will focus on advancing embodied AI and enhancing AI infrastructure efficiency, working in collaboration with university researchers, industry partners, and government agencies. Singapore is positioning itself as a regional AI hub and announced the launch of a testbed on the same day. This platform will assist private companies in co-designing, deploying, testing, and validating commercial AI robotics technology, with industry leaders like Certis, DHL, Grab, and QuikBot expected to be among the first users. Patrick Little of SiFive described Jensen Huang's approach as having "tunnel vision" for where the AI market is headed "five to ten moves" ahead, and acting accordingly. "They have no interest in moving a pawn one step. They want to see the pawn become a queen."
Comments