llms.blog
AI

Ilya Sutskever's SSI Partners with Nvidia, Gains Vera Rubin Access

# Ilya Sutskever's SSI Partners with Nvidia, Gains Access to Vera Rubin Platform Safe Superintelligence Inc., the AI lab founded by former OpenAI chief scientist Ilya Sutskever, has struck a long-term partnership with Nvidia that includes a multi-billion-dollar investment and access to the chipmaker's next-generation Vera Rubin compute platform. The deal, announced July 27, will increase SSI's available compute by an order of magnitude. Nvidia said it decided to invest after gaining rare acces

axc
axc
2 min
Ilya Sutskever's SSI Partners with Nvidia, Gains Vera Rubin Access

# Ilya Sutskever's SSI Partners with Nvidia, Gains Access to Vera Rubin Platform

Safe Superintelligence Inc., the AI lab founded by former OpenAI chief scientist Ilya Sutskever, has struck a long-term partnership with Nvidia that includes a multi-billion-dollar investment and access to the chipmaker's next-generation Vera Rubin compute platform.

The deal, announced July 27, will increase SSI's available compute by an order of magnitude. Nvidia said it decided to invest after gaining rare access to the company's closely guarded research, which it described as having reached "significant research milestones."

"Ilya has pioneered fundamental breakthroughs at the foundation of modern AI, beginning with AlexNet," said Nvidia CEO Jensen Huang. "We are excited to see what new breakthroughs SSI will discover powered by our Vera Rubin platform."

Sutskever, who left OpenAI in May 2024 and founded SSI the following month, said the company has "research that is worthy of scaling up." The partnership gives SSI priority access to Vera Rubin systems, Nvidia's successor to the Blackwell architecture, which began shipping to data center customers in late 2025.

SSI has operated with unusual secrecy since its founding. The company describes itself as "the world's first straight-shot SSI lab, with one goal and one product: a safe superintelligence." It has no public products, no published research papers, and no public benchmarks. Co-founder Daniel Gross, who initially served as CEO, left the company in July 2025 to join Meta's Superintelligence Labs, leaving Sutskever as chief executive.

The company has raised more than $7 billion to date from investors including Andreessen Horowitz, Sequoia Capital, DST Global, and Greenoaks, and was valued at $32 billion prior to the Nvidia deal. Nvidia's additional investment will push that figure higher, though exact terms were not disclosed. The two companies will also collaborate on advancing Nvidia's current and future compute platforms using SSI's research insights.

The partnership places SSI among an elite group of AI labs with guaranteed access to cutting-edge Nvidia hardware at a time when demand for AI compute continues to outstrip supply. Anthropic, OpenAI, and xAI have all signed similarly structured deals with cloud providers and chipmakers in the past year.

Sources: - https://nvidianews.nvidia.com/news/ilya-sutskevers-safe-superintelligence-inc-and-nvidia-announce-long-term-strategic-partnership - https://siliconangle.com/2026/07/27/ilya-sutskevers-safe-superintelligence-gets-access-nvidias-vera-rubin-platform

AILLM
ShareX (Twitter)LinkedIn

Written by

axc

More to read

  • Vector Databases in Production: Comparing Qdrant, Milvus, Weaviate, and pgvector Architecture, Indexing Overhead, Filtered Search, and Serving Economics

    Deploying vector search in production requires navigating fundamental trade-offs across storage topology, indexing latency, memory allocation, and metadata filtering overhead. As retrieval-augmented generation (RAG), multimodal search, and agentic memory architectures scale beyond tens of millions of embeddings, database selection determines whether inference latency remains bounded or collapses under complex filtering constraints. The current vector infrastructure ecosystem divides into two pr

    1 min

  • Activation-Aware Weight Quantization (AWQ): Mathematical Foundations, Salient Weight Protection, and Hardware-Efficient Low-Bit Inference

    Large language model inference during autoregressive generation is overwhelmingly memory bandwidth bound. While the prefill phase processes multiple prompt tokens in parallel with high arithmetic intensity, the token generation phase computes matrix-vector multiplications ($M=1$) for each sequential token. In this regime, the GPU spends the vast majority of its cycle budget streaming model parameters from High Bandwidth Memory (HBM) or GDDR into SRAM rather than performing floating-point arithme

    1 min

  • Intel Details 256-Core Xeon 7 and 480GB Crescent Island Inference GPU at Hot Chips 2026

    At the Hot Chips 2026 conference, Intel outlined architectural disclosures for three upcoming computing platforms tailored for AI workflows: the Xeon 7 enterprise processor (codename Diamond Rapids), the Crescent Island data center inference GPU, and the Wildcat Lake client processor (Intel Core Series 3). The announcements detail Intel's shift toward modular multi-die packaging, open chiplet interconnect standards, and expanded on-chip memory to meet the computational demands of multi-agent AI

    1 min