NVIDIA's blueprint for the future at CES

Source: NVIDIA. Huang speaks at CES 2026.

At CES 2026, NVIDIA founder and CEO Jensen Huang unveiled Rubin, NVIDIA’s first extreme-codesigned, six-chip AI platform now in full production, and introduced Alpamayo, an open reasoning model family for autonomous vehicle development — part of a push to bring AI into every domain.

With Rubin, NVIDIA aims to “push AI to the next frontier” while slashing the cost of generating tokens to roughly one-tenth that of the previous platform, Huang said, making large-scale AI far more economical to deploy. AI services are delivered through the consumption of tokens.

Named after pioneering American astronomer Vera Rubin, the NVIDIA Rubin platform succeeds NVIDIA Blackwell architecture and is the company’s first extreme-codesigned, six‑chip AI platform. Extreme codesign — designing all components together — is essential because scaling AI to gigascale requires tightly integrated innovation across chips, trays, racks, networking, storage and software to eliminate bottlenecks and dramatically reduce the costs of training and inference, Huang explained.

Rubin platform components span:

• Rubin GPUs with 50 petaflops of NVFP4 inference. NVFP4 is a 4-bit floating-point data format that NVIDIA introduced with its Blackwell architecture
• Vera CPUs engineered for data movement and agentic processing
• NVLink 6 scale‑up networking
• Spectrum‑X Ethernet Photonics scale‑out networking
• ConnectX‑9 SuperNICs. NIC stands for network interface card
• BlueField‑4 DPUs or data processing units

Alpamayo, on the other hand, comprises an open portfolio of reasoning vision language action models, simulation blueprints and datasets enabling level 4‑capable driving autonomy. This includes:

• Alpamayo R1 — the first open, reasoning vision-language-action (VLA) model for autonomous driving
• AlpaSim — a fully open simulation blueprint for high‑fidelity autonomous vehicle (AV) testing 

“Not only does it take sensor input and activates steering wheel, brakes and acceleration, it also reasons about what action it is about to take,” Huang said, disclosing that the first passenger car featuring Alpamayo built on the NVIDIA DRIVE full-stack autonomous vehicle platform to be on US roads will be the all‑new Mercedes‑Benz CLA.

“Our vision is that, someday, every single car, every single truck will be autonomous, and we’re working toward that future,” Huang said. 

Huang further introduced the NVIDIA Inference Context Memory Storage Platform — AI‑native storage. The KV‑cache tier that boosts long‑context inference does so with 5x higher tokens per second, 5x better performance per TCO dollar and 5x better power efficiency, translating to cheaper tokens. KV, a way to optimise memory, stands for 'key value'. 

Huang brought all the seemingly-disparate announcements together by explaining that NVIDIA builds entire systems because it takes a full and optimised stack to deliver AI breakthroughs. 

“Our job is to create the entire stack so that all of you can create incredible applications for the rest of the world,” he said.

He emphasised the role of NVIDIA open models in this process. As the domain-specific models are trained on NVIDIA supercomputers, a global ecosystem of intelligence is generated as a foundation for developers and enterprises to build on.

“Now on top of this platform, NVIDIA is a frontier AI model builder, and we build it in a very special way. We build it completely in the open so that we can enable every company, every industry, every country, to be part of this AI revolution,” Huang shared.

The open model portfolio spans six domains, with new, smarter models appearing every six months: Clara for healthcare, Earth-2 for climate science, Nemotron for reasoning and multimodal AI, Cosmos for robotics and simulation, GR00T for embodied intelligence and Alpamayo for autonomous driving.

“These models are open to the world,” Huang said, underscoring NVIDIA’s role as a frontier AI builder with world-class models topping leaderboards. 

“You can create the model, evaluate it, guardrail it and deploy it.”

At CES, NVIDIA also announced that DGX Spark delivers up to 2.6x performance for large models, with new support for the Lightricks LTX‑2 and FLUX image models, as well as upcoming NVIDIA AI Enterprise availability.

“The faster you train AI models, the faster you can get the next frontier out to the world,” Huang invited.

“This is your time to market. This is technology leadership.”

Vendors integrating NVIDIA AI to power their products include Palantir, ServiceNow, Snowflake, CodeRabbit, CrowdStrike, NetApp and Semantec.

“Whether it’s Palantir or ServiceNow or Snowflake — and many other companies that we’re working with — the agentic system is the interface,” Huang shared.

Explore

Watch the keynote video at https://youtu.be/vyzHPVufFbk?si=Z_JhU6cUwKJ4-8Uy