Our Blog

Blog Index

NVIDIA Invests $5 Billion in Ilya Sutskever's Safe Superintelligence: A Strategic Bet on Vera Rubin Compute

Posted on 29th Jul 2026 06:04:37 in Artificial Intelligence, Machine Learning

Tagged as: NVIDIA, Safe Superintelligence, SSI, Ilya Sutskever, AI Investment, Vera Rubin, GPU, Artificial Intelligence, AI Research

In one of the largest single strategic investments of the ongoing AI boom, NVIDIA has committed $5 billion to Safe Superintelligence Inc. (SSI), the AI research lab founded by former OpenAI co-founder and chief scientist Ilya Sutskever. The deal, announced on July 27, 2026, pairs a substantial equity investment with a long-term compute partnership that grants SSI priority access to NVIDIA's next-generation Vera Rubin GPU platform — effectively increasing the startup's compute capacity by an order of magnitude.

The partnership marks a major milestone for SSI, which has operated in near-total stealth since its founding in June 2024. After two years of quiet research with no public product releases or revenue disclosures, SSI has signaled that its internal work has reached a stage ready for massive scale-up.

Inside the NVIDIA-SSI Partnership

The deal combines two components that have become hallmarks of NVIDIA's strategy with frontier AI labs: a direct equity investment and a multiyear compute supply agreement. Under the terms, NVIDIA has made a $5 billion equity investment in SSI, as reported by Bloomberg and confirmed by Reuters from sources familiar with the transaction. Simultaneously, SSI gains access to NVIDIA's upcoming Vera Rubin architecture, a next-generation GPU platform expected to power the largest AI training clusters in the world.

According to the joint statement from both companies, the Vera Rubin access will enable SSI to "increase its compute by an order of magnitude" — roughly ten times its current capacity. The two companies will also collaborate on the technical advancement of NVIDIA's current and future compute platforms, leveraging SSI's research insights into the future of AI.

The deal structure mirrors NVIDIA's pattern with other frontier labs. In March 2026, NVIDIA made a similar significant investment in Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, alongside providing access to the Vera Rubin platform. That strategic template — capital bundled with guaranteed compute — has become NVIDIA's primary method for staying embedded with the labs most likely to produce the next architectural breakthroughs in AI.

NVIDIA was already an investor in SSI prior to this round. The chipmaker said it entered the expanded partnership after obtaining "rare access into the company's closely guarded research," indicating that SSI had shared enough of its internal progress to justify one of NVIDIA's largest-ever funding commitments.

What Is Safe Superintelligence?

Safe Superintelligence Inc. was founded in June 2024 by Ilya Sutskever, Daniel Gross, and Daniel Levy. The company describes itself as "the world's first straight-shot SSI lab, with one goal and one product: a safe superintelligence." Unlike virtually every other major AI company, SSI has released no commercial products, published no public benchmarks, and generates zero revenue. Its entire thesis rests on the conviction that achieving safe artificial superintelligence requires a dedicated, distraction-free research effort unencumbered by short-term commercial pressures.

Ilya Sutskever brings extraordinary credibility to this mission. His research contributions span some of the most significant breakthroughs in modern AI history: AlexNet, which ignited the deep learning revolution; AlphaGo, which demonstrated reinforcement learning at superhuman levels; Sequence-to-Sequence learning, which transformed machine translation; and the GPT family of models that underpin today's generative AI boom. Most recently, Sutskever spearheaded the research that led to reasoning models such as OpenAI o1. His stature in the field is a primary reason SSI has attracted billions in funding without releasing a product.

Daniel Levy, SSI's other co-founder, previously served as a member of technical staff at OpenAI. Co-founder Daniel Gross left the company in 2025 for Meta. The company maintains offices in Palo Alto, California, and Tel Aviv, Israel, with fewer than 100 employees — giving it one of the highest per-employee valuations of any venture-backed company in history.

SSI's Funding History and Valuation

SSI's fundraising trajectory has been remarkably steep for a company with zero revenue:

  • September 2024: $1 billion seed round at a $5 billion valuation, backed by Sequoia Capital, Andreessen Horowitz, DST Global, and SV Angel.
  • April 2025: $2 billion raise at a $32 billion valuation, led by Greenoaks Capital, with Alphabet and NVIDIA joining as strategic investors. Alphabet's stake came alongside Google Cloud becoming a primary compute supplier.
  • July 2026: $5 billion strategic investment from NVIDIA, bundled with the Vera Rubin compute agreement. The investment was structured at or near the existing $32 billion valuation rather than as a fresh up-round.

In total, SSI has raised approximately $7 billion since its founding. The decision to keep the valuation flat at $32 billion across the latest $5 billion addition suggests that even SSI's own backers are taking a measured approach — pouring in capital at the existing price rather than pushing for a higher mark without a product to justify it.

The company's broader investor syndicate also includes Lightspeed Venture Partners, GV, and other top-tier venture firms alongside the anchor investors.

Why NVIDIA Is Betting Big on a Pre-Revenue Lab

NVIDIA's $5 billion investment can be understood as a strategic chess move rather than a conventional financial bet. The chipmaker has navigated the AI boom by recognizing that the scarcest resource in frontier AI research is not capital or talent — it is access to cutting-edge compute hardware. By locking SSI into a multiyear Vera Rubin supply agreement, NVIDIA secures a marquee customer for its most advanced products while simultaneously collecting a share of whatever value SSI's research ultimately produces.

This creates what analysts call a "circular" deal structure — NVIDIA is simultaneously SSI's shareholder and its primary supplier. The same pattern has characterized most large AI infrastructure deals of the current cycle, where capital and compute are essentially the same transaction.

For SSI, the deal solves the single biggest constraint on any frontier-model research effort: guaranteed access to scarce compute. AI training clusters powered by next-generation GPUs are increasingly difficult to procure, with months-long waiting lists and multi-billion-dollar commitments. By securing Vera Rubin capacity through the NVIDIA partnership, SSI sidesteps what would otherwise be a critical bottleneck.

"We have research that is worthy of scaling up, and having access to a big NVIDIA computer will let us do so," Sutskever said in the joint announcement. "We're incredibly proud to be partnering with Jensen and the NVIDIA team, and we are confident that our big bet on the Vera Rubin platform will take us to the next level."

NVIDIA CEO Jensen Huang emphasized Sutskever's research legacy: "Ilya has pioneered fundamental breakthroughs at the foundation of modern AI, beginning with AlexNet. We are excited to see what new breakthroughs SSI will discover powered by our Vera Rubin platform."

Industry Context and Implications

The NVIDIA-SSI deal comes amid a broader wave of chipmaker investments in AI research labs. The pattern reflects a fundamental shift in how the AI industry finances itself: capital and compute are increasingly inseparable, with chipmakers like NVIDIA and AMD deploying strategic investments to lock in demand for their most advanced products.

Beyond the NVIDIA deal, the landscape of frontier AI funding in 2026 has been dominated by billion-dollar-plus rounds. Anthropic raised a $65 billion Series H in June at a $965 billion valuation and is reportedly preparing for an IPO. OpenAI launched its GPT-5.6 family to general availability in July, with the Sol, Terra, and Luna variants becoming the default models across ChatGPT and the API. Google shipped Gemini 3.6 Flash on July 21, and xAI took Grok 4.5 public on July 8.

What sets SSI apart from every other major AI lab is its complete lack of commercial output. While Anthropic sells Claude subscriptions, OpenAI charges for ChatGPT and API access, and Google integrates Gemini across its ecosystem, SSI has nothing on the market and has disclosed no timeline for releasing anything. Its $32 billion valuation is a pure bet on the likelihood that Sutskever and his team will produce a breakthrough that justifies the investment — a bet that NVIDIA has now doubled down on with $5 billion of its own capital.

The partnership also signals that NVIDIA is willing to write increasingly large checks to shape the direction of AI research. With SSI already sitting on what the chipmaker describes as "closely guarded research" breakthroughs, the Vera Rubin-powered compute expansion could accelerate whatever timeline the lab has set for its first public demonstration — potentially producing one of the most significant moments in the field since the launch of ChatGPT.

Looking Ahead

For SSI, the NVIDIA partnership removes the compute bottleneck that constrains virtually every AI research lab. For NVIDIA, it ensures that one of the most secretive and potentially impactful research efforts in AI runs entirely on its silicon. Both sides have signaled that the arrangement is foundational rather than transactional, with Sutskever describing it as a "big bet on the Vera Rubin platform" and Huang framing it as a bet on Sutskever's ability to deliver another fundamental breakthrough.

The coming months will reveal whether SSI's research — guarded so tightly that even its $7 billion in investors have only glimpsed it — lives up to the extraordinary bet NVIDIA has placed on it. For an industry accustomed to rapid product cycles and quarterly metrics, SSI represents a radically different model: a well-funded, commercially silent research lab asking the world to wait and see what emerges.

Sources

Recent Posts

Catgories

Archives

whatsapp me