THE_COMPUTE

company · Semiconductors

NVIDIANVDA

The company the entire buildout orbits: its GPUs (Hopper → Blackwell → Vera Rubin) are the unit of account for AI capacity, its data center segment books over $100B a year, and its CUDA software moat converts silicon share into ecosystem lock-in. Rack-scale NVL systems made liquid cooling mandatory and reset data center power density.

nvidia.com ↗View in relationship graph →

Chart via TradingView. Not investment advice.

Claims on record

  • Each NVIDIA GB300 AI server rack in the SpaceX-Foxconn order costs around $4 million

    UNVERIFIED · Wccftech · Jul 19, 2026

  • The Japanese AI factory will house NVIDIA Vera Rubin NVL72 racks and scale through Spectrum-X networking.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • Jensen Huang pledged on July 16, 2026 to help Japan build its AI-powered future during a Japan visit.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • NVIDIA's Vera CPUs have entered volume production.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • Jensen Huang stated AI infrastructure needs to be built for at least a decade to support long-term demand.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • Jensen Huang stated 'we are a long way from an AI bubble.'

    UNVERIFIED · Wccftech · Jul 17, 2026

  • NVIDIA stated the AI factory will play a role in advancing a global AI market representing a $133 billion TAM by 2040.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • The Japanese AI factory will be able to output 140MW of data center capacity based on the NVIDIA DSX platform.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • The Japanese Vera Rubin AI factory will consist of 13,750 Vera CPUs and 27,500 Rubin GPUs.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • NVIDIA announced Japan's first Vera Rubin AI factory, in partnership with Noetra Corp as part of Japan's FRONTia Project.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • NVIDIA CEO Jensen Huang stated the Vera Rubin platform is in production with 'giant amounts' of production incoming, denying chip delay rumors.

    UNVERIFIED · Wccftech · Jul 17, 2026

  • The review sample of the Lenovo ThinkStation P3 Ultra SFF Gen 2 came with an NVIDIA RTX 4000 Ada Generation SFF card with 20GB GDDR6

    UNVERIFIED · ServeTheHome · Jul 14, 2026

  • NVIDIA's B30A chip could have more than 10x the FLOPs of the H20, and be positioned at half the price and half the performance of the B300.

    UNVERIFIED · SemiAnalysis

  • The US government indicated NVIDIA will be granted a license to export the H20 chip to China after previously banning it.

    UNVERIFIED · SemiAnalysis

  • NVIDIA purchased $5 billion worth of Intel stock, with Intel shares rising 25%

    UNVERIFIED · TechPowerUp

  • The NVIDIA H200 NVL is based on the Hopper generation architecture.

    UNVERIFIED · ServeTheHome · Jul 17, 2026

  • The NVIDIA RTX Pro 6000 Blackwell Server Edition GPU has 96GB of GDDR7 memory capacity.

    UNVERIFIED · ServeTheHome · Jul 17, 2026

  • The NVIDIA H200 NVL has approximately 144GB of memory capacity.

    UNVERIFIED · ServeTheHome · Jul 17, 2026

  • A pre-worn, signed leather jacket worn by NVIDIA CEO Jensen Huang sold at auction at Sotheby's for $960,000.

    UNVERIFIED · CNBC · Jul 17, 2026

  • 3 GB GDDR7 memory modules are being used in NVIDIA's AI GPU products including the RTX PRO 6000 Blackwell and Rubin CPX

    UNVERIFIED · TechPowerUp

  • The RTX 5070 Ti SUPER and RTX 5080 SUPER were each expected to offer 24 GB of GDDR7 memory

    UNVERIFIED · TechPowerUp

  • The RTX 5070 SUPER was expected to offer 18 GB of GDDR7 memory

    UNVERIFIED · TechPowerUp

  • The RTX 50 SUPER series was rumored to use denser GDDR7 memory modules with 3 GB capacity per chip

    UNVERIFIED · TechPowerUp

  • The RTX 50 SUPER launch is on hold due to a shortage of GDDR7 memory chips

    UNVERIFIED · TechPowerUp

  • NVIDIA has shipped GeForce RTX 50 SUPER silicon to its board partners

    UNVERIFIED · TechPowerUp

  • The RTX 5070 SUPER is rated at a TGP of 275 W, compared to 250 W for the original RTX 5070

    UNVERIFIED · TechPowerUp

  • The RTX 5070 Ti SUPER is rated at a TGP of 350 W, a 50 W increase over the RTX 5070 Ti's 300 W

    UNVERIFIED · TechPowerUp

  • The RTX 5080 SUPER is rated at a TGP of 415 W, a 55 W increase over the RTX 5080's 360 W

    UNVERIFIED · TechPowerUp

  • NVIDIA data center segment revenue was $115.2 billion in fiscal year 2025.

    VERIFIED · SEC EDGAR · Feb 26, 2025

Coverage