How CES 2026 Showcased the Transformative Impact of AI: Key AI Announcements from This Year’s Event

This year’s Consumer Electronics Show (CES) marked a pivotal transformation in focus, shifting prominently towards artificial intelligence (AI) advancements instead of traditional consumer product showcases. While this change may have left many gamers feeling underwhelmed, the overwhelming influence of AI in today’s market compelled major tech companies like NVIDIA to represent this trend at the event.

The demonstrations presented by NVIDIA, AMD, and others made it clear that AI is advancing into a new era, moving from merely training models to practical applications involving both physical and agentic AI technologies. Significant computing capabilities are essential for this evolution, resulting in a dominant emphasis on AI infrastructure in keynote presentations delivered by industry leaders, such as Jensen Huang from NVIDIA and Lisa Su from AMD.

In this article, we will highlight some of the most impressive AI innovations showcased at CES 2026, organizing them into various categories for clarity and depth.

NVIDIA’s Vera Rubin: Best AI Platform

NVIDIA’s Vera Rubin AI lineup was among the most eagerly anticipated launches of the year. Having first been revealed at GTC 2025, the Vera Rubin series offers a complete overhaul of hardware components, including networking and chip technologies. During the CES 2026 keynote, Jensen Huang provided an in-depth overview of several upcoming Rubin products, which include:

Rubin GPU with 336 billion transistors
Vera CPU featuring 227 billion transistors
NVLINK 6 Switch for enhanced interconnectivity
CX9 and BF4 for advanced networking solutions
Spectrum-X 102.4T CPO for silicon photonics applications

The reason Vera Rubin garnered our “Best AI Platform”recognition is not solely due to technological advancements, but also NVIDIA’s remarkable ability to transition this architecture into “full production”within a mere nine-month timeline—covering critical stages such as tape-outs and validations. For context, NVIDIA’s Blackwell Ultra began mass production in Q3 2025, with shipments starting in Q4 2025, illustrating the profound disruption NVIDIA is creating in the AI sector.

A person holding two NVIDIA Blackwell chips, labeled 'NVIDIA H100 T1' and 'NVIDIA B100 2025 0242A2M00'

A comprehensive technical analysis of Vera Rubin is available, detailing the configuration of the Rubin GPU, Vera CPUs, and their integration in the NVL72 rack. At CES, NVIDIA presented performance metrics showcasing that the Rubin architecture provides a remarkable 5x improvement in NVFP4 inference compared to Blackwell, along with factory throughput surging up to 10 times better than the GB200, especially within Mixture of Experts (MoE) frameworks.

A presentation slide comparing NVIDIA's Blackwell NVL72 and Rubin NVL72, indicating reduced GPU requirements and increased tokens in factory throughput.

The Vera Rubin AI platform signifies a significant leap in AI infrastructure as we approach 2026. Anticipated customer shipments are set to commence in the second half of the year, potentially allowing hyperscalers to initiate training on frontier models using Rubin before 2026 concludes, a truly remarkable prospect.

AMD’s Helios: Best AI Rack

In discussing AI racks, the focus extends beyond sheer performance to encompass innovative computing arrangements, networking designs, and supporting features. AMD unveiled its Helios rack at CES 2026, built on the Meta Open Rack Wide (ORW) specification, which is critical for enhancing data center efficiency.

AI Compute Performance: 2.9 Exaflops
Memory: 31 TB HBM4
Scale-Out Bandwidth: 43 TB/s
Manufacturing Process: Advanced 2nm / 3nm
CPU Cores: 4, 600 “Zen 6″CPU cores

The Helios rack’s 21-inch internal width allows AMD to fit its larger Instinct MI455X trays efficiently. The ORW specification is tailored for optimal utilization of data center floor space, incorporating a “double-wide”physical frame to facilitate effective laminar airflow. Furthermore, a centralized 48V DC busbus along the back powers individual compute nodes, enhancing operational efficiency.

An AMD server labeled Helios displayed prominently against a blue backdrop. — Image Credits: AMD

A significant advantage of the Helios rack, based on the ORW, is its modifiable design, allowing seamless integration into existing data center configurations. With the implementation of UALink and ORW-compliant designs, AMD aims to alleviate the concerns hyperscalers have about adopting new rack systems. Unlike NVIDIA’s integrated single-unit design, AMD’s Helios offers easier integration into operational ecosystems.

A labeled diagram of the HELIOS server cabinet demonstrating cable management and compute unit layout. — Image Credits: Wccftech/AI

With compelling features and the next-generation Instinct MI400 series AI chips, AMD’s Helios intends to strengthen its competitive edge in AI infrastructure, challenging NVIDIA’s stronghold in the market.

NVIDIA’s Alpamayo: Best AI Software Showcase

In a strategic shift, NVIDIA emphasized its contributions to the open-source model movement at CES 2026, positioning itself ahead of Chinese AI leaders like DeepSeek and Baidu. The company’s Nemotron offerings have emerged as frontrunners in the open-source domain. At CES, NVIDIA unveiled its plan to apply its expertise with open models to autonomous vehicles, showcasing “Alpamayo.”As described by NVIDIA:

NVIDIA Alpamayo is an open portfolio of AI models, simulation frameworks, and physical AI datasets designed to accelerate the development of safe, transparent, and reasoning-based autonomous vehicles. Built for Level 4 autonomy, Alpamayo lets vehicles perceive, reason, and act with human-like judgment, while providing the interpretability and openness required for safety validation and regulatory collaboration.

An NVIDIA display depicts a street scene with pedestrian interaction and trajectory prediction graph overlay. — Image Credits: NVIDIA

NVIDIA seeks to democratize vehicle autonomy through Alpamayo by providing developers with fully open models, simulation frameworks, and datasets tailored for customization in compliance with regulatory standards. Notably, its reasoning-based approach has fast-tracked Level 4 driving capabilities, which are predominantly seen in companies like Waymo and Mercedes-Benz.

A presentation slide detailing components crucial for autonomous vehicle operation. — Image Credits: NVIDIA/Yahoo Finance

The unveiling of Alpamayo demonstrates NVIDIA’s commitment to transforming the AI landscape, particularly in autonomous vehicle technology. By making its resources open-source, NVIDIA enhances the potential for widespread use and integration across various developer frameworks.

AMD’s Ryzen AI Halo: Best Edge AI Platform

A person holding an AMD device with a textured pattern. — Image Credits: AMD

While NVIDIA’s DGX Spark emerged as a vital product at Computex 2025 for its inclusive approach to computing accessibility, AMD is early implementing a similar vision with Ryzen AI Halo. The company plans to introduce full support for the revamped ROCm 7.2.2 suite, optimizing for developer-friendly applications like LM Studio, ComfyUI, and VS Code.

An exploded view of an AMD gaming console showing its components. — Image Credits: AMD

However, with devices like Ryzen AI Halo, pricing remains a critical consideration. NVIDIA’s DGX Spark typically retails for around $4, 000 for high-end configurations, creating affordability barriers for average developers. While AMD hasn’t disclosed pricing for the Ryzen AI Halo, expectations are high that it will be competitively priced, possibly offering a more accessible alternative to NVIDIA’s establishment.

Conclusion: AI, AI & AI at CES 2026

Despite CES’s heritage as a consumer electronics showcase, this year’s event underscored the undeniable significance of AI technology, showcasing how it continues to reshape the landscape for consumers, amidst rising costs and persistent memory shortages. The advancements featured at CES 2026 provide a promising glimpse into the future of AI, especially as emerging sectors like physical and agentic AI begin to take root and flourish.

Source & Images