Nvidia: Star Attraction at CES 2026

2026-02-19 · Source: Big Data & AI News - EE Times · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Cloud Computing & IT Infrastructure · Depth: Advanced, long

Summary

Nvidia unveiled its Vera Rubin platform at CES 2026, featuring six new co-designed chips aimed at significantly boosting performance for large AI systems and AI factories. Key components include the Vera CPU, Rubin GPU, NVLink 6 Switch, ConnectX-9, BlueField-4 DPU, and Spectrum-6 Ethernet Switch. The Rubin GPU, with 336 billion transistors and HBM4 memory, offers 2x to 5x performance improvement over Blackwell. The platform scales from individual chips to DGX SuperPODs, which combine eight NVL72 racks, each containing 72 Rubin GPUs and 36 Vera CPUs, totaling 220 billion transistors per NVL72 rack. Nvidia also introduced the Alpamayo AV platform, an AI model with vision language action (VLA) reasoning and a physical AI open dataset, alongside an expanded software portfolio and a robust AV ecosystem.

Key takeaway

For CTOs and AI infrastructure architects planning large-scale AI deployments, Nvidia's Vera Rubin platform offers a comprehensive, co-designed hardware and software stack. Your decision to invest in this integrated ecosystem could simplify scaling, enhance performance predictability, and improve security for AI factories. Consider evaluating the DGX SuperPOD as a foundational unit for your AI training and inference needs, leveraging its rack-scale coherence and software integration.

Key insights

Nvidia's Vera Rubin platform integrates co-designed chips and software to build scalable AI factories.

Principles

• Co-design chips for system-level optimization.
• Treat the data center as the unit of compute.
• Scale AI systems from chips to racks and SuperPODs.

Method

Nvidia's approach involves co-designing six specialized chips (CPU, GPU, network, DPU) to form modular subsystems (Superchip, Compute Tray, Switch Tray) that scale into NVL72 racks and DGX SuperPODs, creating AI factories.

In practice

• Utilize Rubin GPUs for continuous AI training and inference.
• Implement NVLink 6 for high GPU-to-GPU bandwidth.
• Leverage BlueField-4 DPUs for secure AI factory operations.

Topics

• NVIDIA Vera Rubin Platform
• AI Factories
• Autonomous Vehicles
• Rubin GPU
• Physical AI

Best for: Computer Vision Engineer, Investor, CTO, AI Architect, Machine Learning Engineer, MLOps Engineer

Related on AIssential

• The NVIDIA Rubin Platform: Six New Chips, One AI Supercomputer — The Vera Rubin architecture integrates co-designed CPUs, GPUs, and DPUs for unprecedented AI supercomputing performance and scalability.
• Inside the NVIDIA Rubin Platform: Six New Chips, One AI Supercomputer — NVIDIA's Rubin platform uses extreme co-design to optimize entire data centers for continuous, scalable AI intelligence production.
• The Rubin Era: How NVIDIA’s New Platform Rewrites the Rules for MoE and Agentic AI — NVIDIA's Vera Rubin platform redefines AI hardware by co-designing six chips for unprecedented performance in MoE and agentic AI.
• NVIDIA GTC Taipei at COMPUTEX: Live Updates on What’s Next in AI — NVIDIA's latest platforms advance AI, robotics, and autonomous vehicles with significant performance and efficiency gains.
• CES 2026: Nvidia promises five times the AI performance and ten times cheaper inference with Vera Rubin — Nvidia's new Vera Rubin platform and Alpamayo software aim to dominate the AI and autonomous driving value chains.
• How the NVIDIA Vera Rubin Platform is Solving Agentic AI’s Scale-Up Problem — Extreme co-design of hardware and software is crucial for economically scaling agentic AI workloads requiring low-latency and high-throughput.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Big Data & AI News - EE Times.