ClaimsJensen Huang

“This incredible body of work now becomes flexible and open for anybody to integrate into,”

Jensen Huang, NVIDIA CEO, keynote speaking 19 May 2025 NVIDIA CEO Envisions AI Infrastructure Industry Worth 'Trillions of Dollars' (NVIDIA Blog) archived copy

Factual AI compute, hardware and data

NVIDIA's NVLink Fusion work is now flexible and open for anybody to integrate into.

Mostly false False in substance, right in a detail.

Factual: Statements about the present or the past, checked against the record: did they hold up?

Said for an organisation they lead or work for; it counts on both records.

What would count as true or false

Judged as of 2025-05-19. True: as of the Computex 2025 announcement, NVIDIA made NVLink Fusion available for third parties (other CPU makers, custom ASIC/accelerator designers, hyperscalers) to integrate their chips with NVIDIA's NVLink fabric, with licensing or IP/chiplet access offered to outside companies generally. Mostly true: available to third parties but through partner programs or commercial licensing, not to literally anyone, and announced as available while some integrations were still coming. Mixed: open only to a named, limited set of partners, or announced with availability still pending. Mostly false: integration was restricted (e.g., every configuration had to include NVIDIA GPUs or CPUs) so that 'open for anybody' substantially overstates it, or access was only a future plan. False: NVLink Fusion was not offered to third parties at all.

Set before the research began.

Evidence (4)

  1. E1 NVIDIA Unveils NVLink Fusion for Industry to Build Semi-Custom AI Infrastructure With NVIDIA Partner Ecosystem NVIDIA (via GlobeNewswire, investor.nvidia.com) · 18 May 2025 Own record

    “NVLink Fusion opens NVIDIA’s AI platform and rich ecosystem for partners to build specialized AI infrastructures.”

  2. E2 Nvidia launches NVLink Fusion to connect any CPU with its GPU, AI stack Constellation Research · 19 May 2025

    “NVLink Fusion allows cloud providers and presumably sovereign AI efforts and ultimately private infrastructure to use any ASIC or CPU to scale out Nvidia GPUs.”

  3. E3 Integrating Semi-Custom Compute into Rack-Scale Architecture with NVIDIA NVLink Fusion | NVIDIA Technical Blog NVIDIA Technical Blog · 19 May 2025 Own record

    “NVLink Fusion includes a NVLink chip, for scaling up NVIDIA plus semi-custom ASIC hybrid infrastructure using the same NVIDIA rack-scale architecture, paired with NVIDIA CPUs, NVIDIA NVLink Switches, NVIDIA ConnectX Ethernet SuperNICs, NVIDIA BlueField Data Processing Units (DPU), and NVIDIA Quantum and NVIDIA Spectrum-X switches for scale-out solutions.”

  4. E4 Nvidia launches NVLink Fusion to connect any CPU with its GPU, AI stack Constellation Research · 19 May 2025

    “Initially, MediaTek, Marvell, Alchip Technologies, Astera Labs, Synopsys and Cadence are the early adopters of NVLink Fusion for its custom silicon.”

Reasoning

NVIDIA's announcement (E1) describes NVLink Fusion as opening its platform and ecosystem to partners building specialized AI infrastructure. The independent Constellation Research report (E4) names six companies as the initial adopters. Neither of the two integration paths drops NVIDIA hardware: third-party CPUs connect in order to scale out NVIDIA GPUs (E2), and custom ASICs are built into hybrid systems with NVIDIA CPUs, NVLink switches and NVIDIA networking (E3). Because every configuration includes NVIDIA GPUs or CPUs, "open for anybody to integrate into" substantially overstates what was offered, which the criteria rate mostly false.

Cite this claim

Jensen Huang, “NVIDIA's NVLink Fusion work is now flexible and open for anybody to integrate into.” (said 19 May 2025). Verdict: mostly false. Reliability API, claim c_43599e6f82f3, method 0.1.0+03d23c0e, dataset release v1.1.2. https://epistacklabs.com/claims/c_43599e6f82f3

This claim is in dataset release v1.1.2, with the same verdict.

Reply to this claim

For Jensen Huang, or someone speaking for them: a correction or counter-evidence. Replies are read before anything is shown; an approved reply shows your name, role, date, statement and link. Your email address is never shown and is deleted 30 days after the decision (privacy).

More factual claims by Jensen Huang

  1. Mostly true Factual 4 September 2026
    “we invented a technology called a GPU, and basically a GPU does parallel processing tasks”

    NVIDIA invented the GPU, which does parallel processing tasks such as fluid dynamics, particle physics, quantum chemistry, graphics and AI.

  2. Resolved false Factual 1 June 2026
    “Today, NVIDIA’s token cost is the lowest in the world, not by 10%, by X factors, orders of magnitude, all because we did extreme co-design”

    NVIDIA's token cost is the lowest in the world, lower than others by large factors and orders of magnitude, due to extreme co-design.

  3. Mostly true Factual 1 June 2026
    “It’s the first one to be PCI Express Gen 6.”

    Vera is the first CPU to support PCI Express Gen 6.

  4. Mixed Factual 17 March 2026
    “We have demonstrated that NVFP4 can achieve inference without loss of precision, while delivering significant performance and energy efficiency improvements, and it is equally suitable for training.”

    NVIDIA has demonstrated that NVFP4 can achieve inference without loss of precision, with significant performance and energy-efficiency gains, and is equally suitable for training.

  5. Mixed Factual 17 March 2026
    “Over the past two years, the computational demands for inference have increased approximately 10,000 times, while usage has increased approximately 100 times.”

    Over the past two years, computational demand for inference rose about 10,000 times while usage rose about 100 times.

All of them

Verdicts are made automatically; how verdicts are made.