Thursday, June 25, 2026
EN·DarkSubscribe
AI Infrastructure · News & Analysis
HomeChips & HardwareReport
Chips & Hardware · Report

Nvidia rolls out Vera Rubin GPU architecture as next-generation successor to Blackwell

Confirms Nvidia roadmap continuity and sustained architectural advantage in AI compute over 18–24 months
Trade pressSlicast · June 25, 2026 · US · Source: Google News
importance 87

Nvidia has reached a $5 trillion market valuation, a milestone that coincided with the unveiling of its Vera Rubin platform—an architecture positioned as a dramatic leap forward from its Blackwell generation.

Announced on March 16 at Nvidia's GTC developer conference, Vera Rubin delivers 3.5 times faster model training and 5 times faster inference compared to Blackwell. The training improvement is substantial: AI models that previously required weeks to train could now be completed in days. The inference speedup may prove even more consequential, as inference is where trained models perform their actual work—answering questions, generating images, and running autonomous agents.

Full-scale production began ramping up by early June 2026. Nvidia projects approximately $1 trillion in chip demand extending through 2027, positioning the company to address a significant backlog.

The market responded immediately. On March 16, AI-linked cryptocurrency tokens posted sharp gains: FET, NEAR, GRASS, and WLD all gained more than 10%. Two projects emerged as particularly well-positioned to benefit. Bittensor (TAO) operates a decentralized machine learning network where participants contribute computing power—better GPUs translate directly into more capable contributions and a more valuable network. Render (RNDR) applies similar logic to GPU computing, functioning as a decentralized marketplace for rendering and AI workloads.

Nvidia's valuation places it alongside only a handful of companies in history, though material challenges remain. Converting a backlog into revenue requires flawless manufacturing at scale. Supply chain disruptions, geopolitical tensions around semiconductor production, or an unexpected slowdown in enterprise AI spending could all undermine the thesis. For crypto-focused investors, tokens like TAO and RNDR may capture outsized returns relative to Nvidia's stock, but they carry significantly higher risk, less liquidity, and no earnings reports to anchor valuations.

Read the original
Nvidia rolls out Vera Rubin GPU architecture… · Slicast