Nvidia fuels $68.1 B quarter, heralding the “Agentic AI” era with Vera Rubin’s Extreme

Nvidia’s Rubin platform—Rubin GPU, Vera CPU, NVLink 6 Switch, ConnectX‑9, BlueField‑4 and Spectrum‑6—generated $68.1 billion in revenue for the quarter, according to the company’s CES 2026 briefing and reported by Newsletter. The firm said the “Agentic AI” era is now underway, with the new hardware enabling autonomous, self‑optimising AI agents across enterprise workloads.

The announcement highlighted VR NVL72, the second‑generation rack‑scale Oberon architecture, as the core of the Rubin system. Newsletter notes that VR NVL72’s modular design improves integration efficiency and throughput compared with the earlier Grace Blackwell platform, positioning Nvidia ahead of rivals such as AMD’s MI450X Helios racks, Google’s TPU and AWS’s Trainium 3.

Nvidia claims “extreme co‑design” supremacy, integrating compute, networking and storage into a single rack unit. The company asserts it now offers best‑in‑class silicon for every major component—accelerator, switch, NIC and Ethernet fabric—within one AI server design, according to the same Newsletter briefing.

Analysts at Reuters acknowledged the results beat expectations but warned investors will look for higher cash returns. Wall Street’s focus on dividend and buyback policy remains, even as Nvidia’s hardware dominance fuels the broader AI market, per Reuters’ coverage of the earnings call.

The Next Web cautioned that Nvidia’s quarterly performance could set the tone for confidence in the AI hardware sector. If the Rubin platform sustains its momentum, it may cement Nvidia’s role as the de‑facto supplier for next‑generation “Agentic AI” deployments.