Which Companies Has NVIDIA's "Three-Track Investment Architecture" Invested In?

marsbitPublished on 2026-06-04Last updated on 2026-06-04

Abstract

NVIDIA's investment strategy operates through a "three-track architecture," not just its NVentures venture arm. Corporate Development handles massive strategic bets (e.g., $30B in OpenAI, $10B in Anthropic, $20B in Synopsys). NVentures, a small team, focuses on early-stage, financial investments across sectors like quantum computing (Alice & Bob, PsiQuantum), AI infrastructure (OpenRouter, Tensormesh), and biotech. The NVIDIA Inception accelerator provides non-monetary support. This system allows NVIDIA to nurture startups and lock in strategic partners, creating a vast AI ecosystem. This aggressive capital deployment has drawn scrutiny. Critics like Michael Burry and EU regulators question potential "circular financing," where NVIDIA's equity investments in companies (e.g., CoreWeave, OpenAI) facilitate those companies' purchases of NVIDIA hardware, potentially inflating revenue. Supporters view it as a necessary "virtuous cycle" to secure supply and demand in a compute-scarce market. While NVentures' smaller deals appear like traditional VC, its role within the larger, controversial investment framework remains a point of debate.

Author: Ada, Deep Tide TechFlow

Recently, NVentures, NVIDIA's venture capital arm, made a new investment in the French quantum computing company Alice & Bob, focusing on fault-tolerant quantum computing.

It is a common misconception to attribute all of NVIDIA's external investments under the name NVentures. In fact, the collective scale of this venture division's approximately 30 annual investments, established in 2021, pales in comparison to a single investment made casually by the Corporate Development team. The latter's $2 billion equity investment in Synopsys at the end of 2025 alone is several times larger than NVentures' cumulative investment amount over nearly three years.

To understand how NVIDIA uses capital to weave the AI ecosystem, one must start with the "Three-Track Architecture" of its investment system. The Corporate Development team handles strategic large-scale investments and mergers & acquisitions ranging from billions to hundreds of billions of dollars. NVentures is responsible for early-stage, broadly scoped financial investments across industries. NVIDIA Inception is a startup accelerator that provides resource connections without direct capital investment. These three parts work in synergy, forming the largest and fastest-paced capital deployment machine in Silicon Valley's history, and also becoming the core target of "circular financing" allegations in the eyes of short-sellers.

The True Face of NVentures: 2-Person Team, 79 Companies, 20 Unicorns

Although bearing NVIDIA's brand, NVentures' internal size is surprisingly small. According to private equity data firm Tracxn, as of May 2026, the entire team consists of only 2 people, having cumulatively invested in 79 companies, fostering 20 unicorns, including AI video generation platform Synthesia, clinical AI company Abridge, quantum computing company PsiQuantum, and others. In the past 12 months, the team completed 43 new investments, with 20 deals in the first five months of 2026 alone, indicating a significantly accelerated pace.

Leading NVentures is Mohamed "Sid" Siddeek, corporate vice president and head of NVentures. Siddeek's resume itself reflects NVIDIA's positioning for this division. He worked at Morgan Stanley in the late 1990s, accompanying Jensen Huang during NVIDIA's IPO roadshow; then served for nearly a decade as head of TMT & Telecom investments at the UAE sovereign wealth fund Mubadala; later managed enterprise software and healthcare investments at SoftBank's Vision Fund; and finally returned to NVIDIA in 2021 to establish NVentures.

Siddeek's own description of the investment scope is: "The real screening criteria are only two layers: the first is anywhere NVIDIA can reach, the second is which areas are investable." In an interview with Global Corporate Venturing, he revealed this means horizontal coverage across almost all industries AI can transform—healthcare, manufacturing, robotics, autonomous driving, quantum, etc.—and vertical coverage from underlying tools to the application layer, all within NVentures' investment scope.

Three-Track Architecture: Corp Dev for Strategy, NVentures for Early-Stage, Inception for Ecosystem

NVIDIA's external investment system consists of three distinct parts with clear division of labor.

The first layer is the Corporate Development team, led by Vishal Bhagwati, responsible for all strategic-level large-scale investments, joint ventures, and M&A. The magnitude of deals on this track is completely different from NVentures. Representative moves from the second half of 2025 to the first half of 2026 include leading a $30 billion investment in OpenAI in February 2026 (as part of a roughly $110 billion funding round), with a commitment to potentially increase to $100 billion in the future; a $10 billion commitment to Anthropic in November 2025; a $2 billion injection into Synopsys at the end of 2025; a $2 billion follow-on investment in CoreWeave in early 2026, accompanied by a $6.3 billion cloud capacity purchase agreement; a $2 billion investment in Nebius in March 2026; and an equity commitment of up to $2 billion in xAI.

According to CNBC, in just the first four months of 2026, AI-related equity investments led by the Corporate Development team exceeded $40 billion. NVIDIA invested a total of $17.5 billion in private companies and infrastructure funds during fiscal year 2025.

The second layer is NVentures, led by Sid Siddeek, positioned as a traditional venture capital firm pursuing financial returns. Individual deal sizes range from a few million to tens of millions of dollars, primarily investing at Seed to Series B stages. Siddeek explicitly told Global Venturing that NVentures "primarily focuses on early-stage investments, while the Corporate Development team handles larger, more directly strategic investments." Behaviorally, NVentures predominantly participates as a co-investor, leading only about one-eighth of its investments, more often joining rounds led by top VCs like Accel, a16z, and Sequoia, leveraging the NVIDIA endorsement.

The third layer is NVIDIA Inception, essentially a startup accelerator program that does not directly invest capital but provides startups with NVIDIA hardware credits, technical support, market promotion, and VC connection channels. NVIDIA's upgraded "VC Alliance" launched in 2025, partnering with institutions like Accel, Elaia, Partech, and Sofinnova, distributes NVIDIA DGX Cloud Lepton compute credits to their portfolio companies, representing an extension of Inception in Europe.

There is a clear "funnel" relationship among the three. Inception discovers early-stage projects and integrates them into the NVIDIA ecosystem. Those with investment potential come into NVentures' view and may receive early-stage checks from a few million to tens of millions of dollars. When a company grows to a size significant enough to impact NVIDIA's strategic landscape (becoming a major customer, key supplier, or potential acquisition target), it "upgrades" to the Corporate Development team, entering cooperation frameworks worth billions or even hundreds of billions of dollars.

NVentures' Recent Moves: Quantum, Inference Routing, AI Security

NVentures' activity in May 2026 was notable. Just within the recent month, there have been four publicly disclosed deals. On May 22, French quantum computing company Alice & Bob announced NVentures participated in the extension of its €100 million Series B round. Alice & Bob's core technology is a fault-tolerant quantum computing architecture based on "cat qubits," deeply collaborating with NVIDIA's hybrid quantum-classical computing stack like CUDA-Q, cuQuantum, Dynamiqs, and NVQLink. On May 26, AI model routing platform OpenRouter completed a $113 million Series B round, with NVentures co-investing alongside Google CapitalG, Snowflake, and others. OpenRouter's business provides developers a unified interface to access APIs from dozens of different model providers globally. On May 28, AI inference infrastructure startup Tensormesh completed a $20 million seed extension round, with NVentures co-investing alongside CoreWeave, AMD, and others. On May 6, AI cybersecurity company Xbow completed a $35 million Series C extension, with NVentures participating.

Judging by the investment targets, NVentures' recent tilt clearly leans toward three directions: quantum computing (Alice & Bob, Quantinuum, PsiQuantum), AI biopharma (Relation Therapeutics, Genesis Therapeutics), and AI Agent & inference layer (OpenRouter, Tensormesh, etc.). This aligns with Siddeek's statement "anywhere NVIDIA can reach" and also corresponds precisely to the directions NVIDIA is investing in for its next-generation software stacks like CUDA-Q, CUDA-X, and Triton.

Geographically, NVentures' European deployment has notably accelerated. It completed 14 European investments in 2025, double the 7 investments in 2024.

Panoramic View of the Three-Tier Investment Portfolio

If we overlay the portfolios of the three investment tiers on the same map, NVIDIA's "capital radiation" on the AI ecosystem can be summarized into five main quadrants.

The foundation model layer includes OpenAI, Anthropic, xAI, Mistral, Cohere, Thinking Machines Lab, Reflection AI, Black Forest Labs. This layer is primarily funded by the Corporate Development team, with NVentures participating in smaller shares.

The cloud & infrastructure layer includes CoreWeave, Nebius, Lambda, Crusoe, Nscale, Firmus Technologies. This layer is also dominated by the Corporate Development team, with individual investments often reaching billions of dollars, accompanied by long-term compute purchase contracts.

The application & developer tools layer includes Cursor, Perplexity, Synthesia, Runway, Lovable, Together AI, Weka. NVentures has higher participation in this layer, with relatively smaller amounts.

The robotics & autonomous driving layer includes Figure AI (latest valuation $39 billion) and Wayve (valuation $8.6 billion). This involves combined efforts from both Corporate Development and NVentures.

The quantum computing & biopharma layer includes PsiQuantum, Quantinuum, Alice & Bob, Relation Therapeutics. This consists mainly of early-stage investments led by NVentures, serving as NVIDIA's hedging bets on post-GPU-era computing paradigms.

According to venture capital research firm F4 Fund, from 2025 to early 2026, in the investment rounds participated in by NVIDIA (Corporate Development + NVentures), at least 10 companies crossed the $1 billion valuation threshold, including OpenAI, Anthropic, xAI, Mistral, Figure AI, Cursor, Perplexity, Scale AI, Wayve, and others.

Controversy: Burry's Shorts and the Question of "Circular Financing"

However, NVIDIA's massive external investment landscape is attracting increasing scrutiny. The most representative criticism comes from hedge fund manager Michael Burry, famous from the movie "The Big Short."

According to Scion Asset Management's Q3 2025 13F filing, Burry established short positions against NVIDIA and Palantir before September 30, 2025, including put options for approximately 1 million NVIDIA shares, representing a nominal exposure of about $187 million based on the stock price at the time; and 50,000 put option contracts (each for 100 shares) on Palantir, with actual premium payments of about $9.2 million. Burry posted on his X account "Cassandra Unchained" with a screenshot from "The Big Short," captioning "Sometimes, we can see bubbles," and later retweeted a Bloomberg chart about NVIDIA's circular financing, directly pointing the finger at NVIDIA's capital deployment model.

Burry's specific accusations are technical. He estimated in a Substack post that between 2026 and 2028, cloud providers including Microsoft, Google, Oracle, and Meta would cumulatively understate depreciation by about $176 billion by extending the accounting depreciation period for NVIDIA GPUs, thereby inflating profits during that period. This accounting adjustment resonates with NVIDIA's equity investments in customers. The former gives buyers higher "paper profits" to absorb larger capital expenditures, while the latter directly provides buyers with funds to purchase NVIDIA hardware.

At the institutional level, similar suspicions are accumulating. The EU competition regulator explicitly included the "risk of circular spending" within NVIDIA's investment system in its review scope in March 2026. Seaport Research estimates that for every $1 of equity NVIDIA invests, it corresponds to about $3.5 in downstream chip purchase revenue. Bloomberg's March 2026 feature on "AI Circular Trades" mapped the dense web of fund flows among NVIDIA, CoreWeave, OpenAI, Oracle, and Anthropic. NVIDIA holds about a 7% stake in CoreWeave, CoreWeave uses NVIDIA GPUs as collateral for financing, in turn uses cash to purchase more GPUs from NVIDIA, and NVIDIA then signs a $6.3 billion cloud capacity purchase agreement, committing to absorb CoreWeave's excess capacity through 2032. NVIDIA committed up to $100 billion in investment to OpenAI, OpenAI commits to purchasing NVIDIA hardware and building a $300 billion data center via Oracle, and Oracle then purchases GPUs from NVIDIA. NVIDIA invests $10 billion in Anthropic, Anthropic commits to deploying Claude on Microsoft Azure, and Azure purchases NVIDIA Grace Blackwell and Vera Rubin systems.

Counterarguments from supporters also exist. Asset manager Janus Henderson characterized this model as a "virtuous cycle," arguing that in an era of extreme compute scarcity, binding supply and demand through "equity + long-term purchase contracts" is a reasonable commercial arrangement. Morningstar's analysis pointed out that NVIDIA's arrangement with CoreWeave to "commit to purchasing excess capacity" actually makes NVIDIA bear CoreWeave's inventory risk, thus constituting a constraint on the impulse to push hardware sales in the short term.

In this controversy, NVentures' position is rather delicate. Its early-stage, small-scale, co-investment-focused, and industry-dispersed investment style forms a stark contrast to the "circular trade" model of the Corporate Development team. Companies like Alice & Bob, Tensormesh, and OpenRouter, which NVentures invests in, are not large enough to form the "both NVIDIA customer and NVIDIA investee" cycle; their investment behavior is closer to the financial investment logic of a traditional CVC. However, from the perspective of NVIDIA's overall investment system, whether NVentures, to some extent, acts as the "compliant venture capital facade" in external disclosures, making it easier for outsiders to interpret NVIDIA's investment activities as normal venture capital behavior rather than systematic seller financing, is the unspoken but implied question from Burry and EU regulators.

NVIDIA's consistent official stance is that all investments are based on independent business judgment and are not linked to hardware sales. But market observers are increasingly quoting a phrase: in an era of compute shortage, whether to believe "the entanglement of equity and purchase contracts is coincidental" is itself a question of trust.

Related Questions

QWhat are the three parts of NVIDIA's investment system as described in the article?

AThe three-part system consists of: 1) The Corporate Development team, which handles strategic, large-scale investments and acquisitions worth billions to hundreds of billions of dollars. 2) NVentures, the venture capital arm that makes early-stage, financially-focused investments ranging from millions to tens of millions of dollars. 3) NVIDIA Inception, a startup accelerator program that provides resources and connections but does not make direct investments.

QWho leads the NVentures team and what is its typical investment role?

ANVentures is led by Mohamed 'Sid' Siddeek. Its typical investment role is that of a traditional VC seeking financial returns, primarily participating in early-stage rounds (Seed to Series B). It mostly acts as a co-investor in rounds led by top VCs like Accel or Sequoia, with only about one-eighth of its investments being lead roles.

QWhat are some recent investment areas for NVentures as of May 2026?

AAs of May 2026, NVentures's recent investments show a clear tilt towards three areas: Quantum Computing (e.g., Alice & Bob, Quantinuum, PsiQuantum), AI Biopharma (e.g., Relation Therapeutics, Genesis Therapeutics), and AI Agent & Inference Layer (e.g., OpenRouter, Tensormesh).

QWhat is the core criticism from Michael Burry regarding NVIDIA's investment activities?

AMichael Burry's core criticism revolves around the concept of 'circular financing.' He alleges that NVIDIA's equity investments in its customers (like AI companies and cloud providers) effectively provide them with funds to purchase NVIDIA hardware. This, combined with accounting practices that extend GPU depreciation periods for buyers, creates a loop that inflates reported profits and sustains demand for NVIDIA's products.

QAccording to the article, what is the "funnel" relationship between the three parts of NVIDIA's investment system?

AThere is a clear 'funnel' relationship: NVIDIA Inception discovers early-stage projects and brings them into the NVIDIA ecosystem. Promising companies from this pool may then attract NVentures for an early-stage investment. Finally, if a company grows to a size significant enough to impact NVIDIA's strategic landscape (as a key customer, supplier, or potential acquisition target), it 'graduates' to the Corporate Development team for multi-billion dollar strategic partnerships or investments.

Related Reads

The Revelation from the Raydium Theft Incident: New DeFi Vulnerabilities Lurking in Forgotten Old Contracts

**Raydium Exploit Reveals DeFi's Hidden Risk: Forgotten "Zombie" Contracts** A recent attack on Raydium's deprecated V3 AMM pools resulted in a loss of approximately $1.34 million. The hacker exploited pools that were no longer supported by Raydium's current UI or SDK but remained fully functional and accessible on-chain. This incident highlights a critical, often overlooked category of risk in DeFi: inactive or legacy smart contracts that projects fail to properly decommission. Since March 2025, there have been at least 8 publicly reported attacks targeting such abandoned contracts, with total losses around $10.8 million. Including older pools and deprecated features, the count rises to 10 incidents with roughly $22.5 million in losses. These "zombie contracts" represent a lifecycle management failure rather than a code vulnerability, yet they are typically misclassified under general "code bug" categories in security reports, masking the true scale of the problem. The root cause is that projects often merely document a contract as "deprecated" without taking essential technical steps to secure it: withdrawing remaining assets, disabling external call functions, and implementing ongoing monitoring. These forgotten, under-monitored components become prime targets for attackers. To address this, the industry needs to recognize "zombie contracts" as a distinct risk category and establish standardized decommissioning protocols. Essential steps should include: 1) a formal retirement announcement, 2) removal of all front-end integrations, 3) withdrawal of locked assets, 4) disabling key contract functions, 5) ongoing security monitoring, 6) clear user communication, and 7) a post-mortem analysis. The value of a DeFi project lies not only in its current TVL but also in the security of its historical codebase, which has now become a new attack surface.

Foresight News8m ago

The Revelation from the Raydium Theft Incident: New DeFi Vulnerabilities Lurking in Forgotten Old Contracts

Foresight News8m ago

Robots Begin to 'Consume Data': The Hidden Production Chain from Indian Data Factories to Billion-Dollar Humanoid Robots

Robots have started to 'consume data,' driving the formation of a new industrial supply chain focused on producing training data for embodied AI. Unlike large language models, which are trained on vast internet text corpora, embodied AI models face a 'data desert' in the physical world. This has created a massive demand for first-person perspective video data (Ego Data), captured by workers wearing cameras in places like Indian garment factories. Companies like Neocambrian AI are establishing 'data factories' where workers perform standardized tasks (e.g., sorting clothes, kitchen organization) to generate thousands of hours of video. Research, such as NVIDIA's EgoScale, demonstrates that scaling this human demonstration data predictably improves robot performance, particularly for dexterous manipulation. This has validated a training path combining large-scale human data for pre-training with smaller amounts of robot-specific data for fine-tuning. The value of different data types varies significantly, forming a 'data pyramid.' The base consists of low-cost, large-scale internet and Ego Data. Higher layers include more expensive motion-capture data (e.g., from data gloves), simulation/synthetic data, and the most costly and scarce layer: real robot teleoperation data. This demand has spawned a layered ecosystem of data suppliers: low-cost data factories, motion capture and alignment specialists, robot-native teleoperation service providers, simulation data companies, and platforms aiming for data standardization. Robot companies themselves are adopting a 'layered procurement' strategy: outsourcing generic Ego Data while building in-house capabilities for robot-specific adaptation data and the critical deployment/failure data generated in real-world applications. The industry is shifting focus from hardware and basic mobility to the data pipelines required for general-purpose capability. While parallels exist to data labeling companies like Scale AI in the LLM boom, the physical complexity of robot data—involving action success ambiguity and sim-to-real gaps—requires more integrated solutions for data collection, annotation, and a continuous feedback loop. The race is on to build the data engines that will teach robots to operate reliably in the unstructured real world.

marsbit2h ago

Robots Begin to 'Consume Data': The Hidden Production Chain from Indian Data Factories to Billion-Dollar Humanoid Robots

marsbit2h ago

Spicy Commentary | Michael Saylor's 'Player Talk'; 60-Year-Old Aunt Liquidated After 'Scamming a Young Man'

**"Spicy Commentary": Three Tales of Crypto's Wild Week** This week's "Spicy Commentary" column highlights three dramatic stories from the cryptocurrency world. First, **MicroStrategy's Michael Saylor** addressed the controversy over his company potentially selling Bitcoin. At the BTC Prague event, he clarified, "I never said the company can't sell Bitcoin. I told *you* never to sell *your* Bitcoin." This "do as I say, not as I do" stance was criticized by netizens as peak linguistic gymnastics, noting a history of him previously stating the company would "never" sell. Second, a **bizarre fraud case** emerged from Beijing. A 60-year-old woman, obsessed with getting rich from crypto but unwilling to risk her own savings, posed online as the 20-something "god-daughter" of a high-ranking official. She catfished a young man, convincing him to give her over 200,000 yuan for fabricated emergencies. She then invested all the stolen money into cryptocurrency with 10x leverage, only to lose everything in a market crash. The woman was sentenced to four years in prison for fraud. Finally, a **sobering trader's tale** surfaced on Reddit. A user posted "Tale of a crypto trader," confessing their net worth had plummeted from a peak of $45 million to roughly $17,200, primarily due to holding meme coins too long. The post, described as a crypto "book of confessions," sparked reactions ranging from sympathy to critique about greed, poor risk management, and the perils of treating meme coins as long-term investments instead of taking profits. The column concludes that this week featured masterful rhetoric, elaborate scams, and extreme financial volatility, stitching together another chapter in crypto's unpredictable theater.

Foresight News3h ago

Spicy Commentary | Michael Saylor's 'Player Talk'; 60-Year-Old Aunt Liquidated After 'Scamming a Young Man'

Foresight News3h ago

Tremble Humans, AI Continues Its Accelerated Sprint

Trembling, Humans: AI Continues Its Accelerated Sprint Yes, AI is still rapidly accelerating. While deep learning seemed to stall quickly in its early years, large models after years of development show no sign of hitting their ceiling. At the Zhiyuan Conference 2026, the focus is on enabling AI to move from the digital world into the physical world. Scaling Law remains effective, continuing to drive advancements in both large language models and multimodal models. The industry is now entering a phase of pursuing World Models, though unresolved technical paths and data issues mean this exploration may take 3-5 more years. Concurrently, breakthroughs in Agents are accelerating AI's real-world application in fields like healthcare and meetings. Making Agents truly useful requires key hardware-software co-design, evident from the strong presence of chip vendors at the conference. We stand at a new historical threshold where AI is becoming a foundational force reshaping the world. The first day of the conference highlighted AI's evolution from "knowing how to chat" to "knowing how to work." Scaling Law persists, World Models are the next key battleground, and Agents are transitioning from usable to好用 (user-friendly). Scaling Law is not ending but diversifying. New models like Anthropic's Fable 5 demonstrate scaling through parameter size, synthetic data, and reinforcement learning. Advancements in AI Coding and Agent deployment are enabling a trend of AI self-evolution, potentially allowing AI to take over digital world iterations. World Models represent the next frontier for large models extending into the physical realm, but no current model is truly impressive at solving real-world problems. Technical consensus is lacking, with debates on data sources (video, simulation, real-world). Different approaches are emerging: language-centric, pixel-centric, 3D-structure-centric, and visual-representation-centric models. Zhiyuan Institute is exploring a fifth path: unified latent space modeling fusing language and visual representations, and introduced its own under-development World Model, Physis-v0.1. On the product side, Agents are key to bringing AI into daily life. Since 2025, the "Year of the Agent," products have become more proactive and capable of complex tasks. Zhiyuan showcased four vertical Agents for cardiac diagnosis, autonomous research, meeting summarization, and protein risk discovery. However, technical challenges remain, particularly in context engineering like memory and orchestration. "Harness" – the engineering framework around an Agent – is crucial for maximizing its capabilities by clarifying intent, designing workflows, and incorporating validation and feedback. In summary, AI's breakneck pace continues on multiple fronts: foundational model scaling, the ambitious pursuit of World Models for physical understanding, and the ongoing refinement of practical Agents. The journey from capable to truly reliable and useful AI systems is well underway.

marsbit3h ago

Tremble Humans, AI Continues Its Accelerated Sprint

marsbit3h ago

Trading

Spot
Futures

Hot Articles

How to Buy AR

Welcome to HTX.com! We've made purchasing Arweave (AR) simple and convenient. Follow our step-by-step guide to embark on your crypto journey.Step 1: Create Your HTX AccountUse your email or phone number to sign up for a free account on HTX. Experience a hassle-free registration journey and unlock all features.Get My AccountStep 2: Go to Buy Crypto and Choose Your Payment MethodCredit/Debit Card: Use your Visa or Mastercard to buy Arweave (AR) instantly.Balance: Use funds from your HTX account balance to trade seamlessly.Third Parties: We've added popular payment methods such as Google Pay and Apple Pay to enhance convenience.P2P: Trade directly with other users on HTX.Over-the-Counter (OTC): We offer tailor-made services and competitive exchange rates for traders.Step 3: Store Your Arweave (AR)After purchasing your Arweave (AR), store it in your HTX account. Alternatively, you can send it elsewhere via blockchain transfer or use it to trade other cryptocurrencies.Step 4: Trade Arweave (AR)Easily trade Arweave (AR) on HTX's spot market. Simply access your account, select your trading pair, execute your trades, and monitor in real-time. We offer a user-friendly experience for both beginners and seasoned traders.

10.9k Total ViewsPublished 2024.03.29Updated 2026.06.02

How to Buy AR

Discussions

Welcome to the HTX Community. Here, you can stay informed about the latest platform developments and gain access to professional market insights. Users' opinions on the price of AR (AR) are presented below.

活动图片